Practical Computing for Biologists

Release 1.0

Cliburn Chan

June 01, 2012

CONTENTS

Updates

Introduction

Course Description

Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

4.1 Data Samples . . . . . . . . . . . . . . . . . . . . . . .
4.2 Assignments . . . . . . . . . . . . . . . . . . . . . . .
4.3 References . . . . . . . . . . . . . . . . . . . . . . . .
4.4 Participants . . . . . . . . . . . . . . . . . . . . . . . .
4.5 Installation and introduction . . . . . . . . . . . . . . .
4.6 Basic Unix Commands . . . . . . . . . . . . . . . . . .
4.7 Using a text editor and regular expressions . . . . . . .
4.8 Remote computing and web page generation . . . . . .
4.9 Python Basics I . . . . . . . . . . . . . . . . . . . . . .
4.10 Python Basics II . . . . . . . . . . . . . . . . . . . . .
4.11 Python Modules . . . . . . . . . . . . . . . . . . . . .
4.12 NumPy and Matplotlib . . . . . . . . . . . . . . . . . .
4.13 Biopython I . . . . . . . . . . . . . . . . . . . . . . . .
4.14 Biopython II . . . . . . . . . . . . . . . . . . . . . . .
4.15 Data management and relational databases . . . . . . .
4.16 Data analysis with Python . . . . . . . . . . . . . . . .
4.17 Vector graphics with Inkscape . . . . . . . . . . . . . .
4.18 Capstone Example . . . . . . . . . . . . . . . . . . . .

Index

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

7
7
9
10
11
12
13
19
22
30
43
52
57
78
83
85
88
95
98
105

ii

CHAPTER

ONE

UPDATES
24 April 2012 Personal web space on the Duke servers is not turned on by default for DUMC perosnnel. However, if
you make a request for AFS space to [email protected], it will be available to you within 24 hours.
9 April 2012 The PCfB textbook is now available for collection for course participants at Room 120, Surgical Oncology Research Facility. Please read or at least scan the book before the workshop starts. There are also pre-workshop
Assignments that you will need to do. We will shortly be contacting course participants for data sets/repetitive tasks
that could serve as relevant demonstrations or examples of regular expression manipulation, programming or use of
relational databases.
3 April 2012 The course is now fully subscribed, and new registrants will be placed on a wait list. Please continue to
register if you are interested - if there is sufficient demand, we will plan for a second workshop. Thanks so much for
your enthusiasm and support!

Practical Computing for Biologists, Release 1.0

Chapter 1. Updates

CHAPTER

TWO

INTRODUCTION
The CFAR Biostatistics and Computational Biology Core is conducting a free four-day workshop for Duke researchers
to learn how to use the computer more effectively for scientific work. It is designed for people who need to work with
large and complex data sets and suspect that there is a better and faster way to get their work done. The course
will use the textbook Practical Computing for Biologists (PCfB) by Steven Haddock and Casey Dunn, and CFAR is
generously giving each participant a free copy of the book. The main intent of the course is to teach researchers how
to use the Unix shell, the Python programming language, databases and image manipulation tools to execute common
scientific chores. An OS X system is preferred since Macs provide a Unix command line natively. Windows users
can also participate by setting up Linux in an emulator (this is perfectly safe and instructions are given in the PCfB
textbook).
The course is designed for people trained in biology, and no previous Unix or programming experience is necessary.
The course will be limited to 12 participants and will be held at the Surgical Oncology Research Facility (SORF) Beard
Conference Room from 29 May 2012 to 1 June 2012. Please email [email protected] if you have any enquiries
or wish to register for the course. Acceptance will be on a first-come first-serve basis, but CFAR investigators and
their trainees will be given priority.
We will contact course participants before the workshop starts to collect your copy of Practical Computing for Biologists. To make it relevant for your needs, participants will also be asked to suggest computational tasks that you would
like to automate or simplify, as well as to contribute data sets that are tedious to preprocess and filter manually. We
will try to work these examples into the demonstrations or class assignments if at all possible. Updates and course
materials will be posted at http://www.duke.edu/~ccc14/pcfb/.

Practical Computing for Biologists, Release 1.0

Chapter 2. Introduction

CHAPTER

THREE

COURSE DESCRIPTION
29 May 2012 (Tuesday)
AM: Software installation and working with text editors. We will install the TextWrangler editor (jEdit for Linux
users), the Enthought Python distribution (Academic license), ImageMagick, ImageJ, MySQL Community Server and
MySQL Workbench. Participants are expected to install the software ahead of the workshop following instructions in
PCfB, but help and troubleshooting will be provided in the morning session if necessary. Many operations on large
file sets, especially for text data, are performed much more efficiently from the command line than from a graphical
interface. We will learn how to open a Terminal, and perform text processing, access material from the web, and write
simple shell scripts to automate common tasks.
Installation and introduction
Basic Unix commands
PM: We will learn to use the TextWranger/jEdit editor to understand the basics of regular expressions, and how to
reformat text using regular expressions. TextWrangler/jEdit will also be used to develop programs from Day 2. We
will also learn to transfer and synchronize files with remote computers from the command line, or run programs on
remote computers using the command line (ssh)). We will conclude by showing how to construct a simple homepage
using Sphinx and upload it to the Duke server.
Using a text editor and regular expressions
Remote computing and web page generation
30 May 2012 (Wednesday)
AM: Day 2 introduces you to the Python programming language, a modern dynamic language that is (relatively) easy
to learn. The morning session will introduce you to the powerful IPython interpreter, where you will test out code
snippets with instant feedback, and learn about the Python documentation and help system. We wil then move on to
Python scripting, including decisions and loops, reading from and writing to files, and writing your own functions.
Python Basics I
Python Basics II
PM: The afternoon will introduce you to the most useful Python modules in the standard library, followed by an
introduction to the NumPy module for numerical work, and Matplotlib for graphics.
Python Modules
NumPy and Matplotlib
31 May 2012 (Thursday)
AM: You will learn more about Numpy and Matplotlib, together with how to use the Biopython module for sequenc
and array analysis, as well as how to access the NCBI databases programmatically.
Biopython I

Practical Computing for Biologists, Release 1.0

Biopython II
PM: The afternoon starts with an introduction to relational databases and how to query them using SQL, then concludes
with some intermediate examples of using Python for data analysis and statistical simulation.
Data management and relational databasesI
Data analysis with Python
01 June 2012 (Friday)
AM: On the final day, we will have a tutorial for how to create scientific diagrams using the vector illustration program
Inkscape. The course will conclude with working through developing a moderately complex Python program to parse,
summarize and display data from a cytokine assay experiment.
Vector graphics with Inkscape
Capstone example

Chapter 3. Course Description

CHAPTER

FOUR

INSTRUCTOR: CLIBURN CHAN,

BIOSTATISTICS AND
BIOINFORMATICS.
Cliburn is a computational biologist whose main research interest is in data analysis and modeling of immune responses. He teaches the Introduction to the Practice of Biostatistics I & II courses for the Duke Masters in Biostatistics
program, and has been programming in Python for over a decade. Other instructors will be Jacob Frelinger, a PhD
student in the Computational Biology and Bioinformatics (CBB) program and Adam Richards, a postdoctoral fellow
in the department of Biostatistics and Bioinformatics.

4.1 Data Samples

4.1.1 Basic Unix commands
1. hamlet

4.1.2 Using a text editor and regular expressions

1. TextWrangler tutorial
2. Lorem ipsum
3. Email
4. Find and replace
5. Ch3observations

4.1.3 Remote computing and web page generation

No data samples.

4.1.4 Python Basics I

No data samples.

Practical Computing for Biologists, Release 1.0

4.1.5 Python Basics II

1. sequence1
2. hamlet

4.1.6 Python Modules

1. CSV sample data
2. CSV exercise solution

4.1.7 NumPy and Matplotlib

1. cell cycle microarray

4.1.8 Biopython I
1. orchid FASTA file

4.1.9 Biopython II
4.1.10 Data management and relational database
SQLite example database Code to generate the database

4.1.11 Data analysis with Python

1. Ch3observations

4.1.12 Vector graphics with Inkscape

1. tux image

4.1.13 Capstone Example

1. Cytokine assay (Excel)
2. Cytokine assay (TDL)

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

4.2 Assignments
4.2.1 Pre-workshop
#1 Software installation
Once you have collected your copy of the PCfB book from SORF, install the following software. If you will be using
a Windows system, please follow the instructions starting on page 458 under Installing VirtualBox till the end of
Appendix 1.
For Mac users, install TextWrangler (Page 12) and MySQL (Page 260). We also recommend installing the Enthought
Python Distribution by requesting a free academic copy from http://www.enthought.com/products/edudownload.php
(this will email you a download link). It will also be useful to learn how to compile and install software from source
by following the instructions given in Chapter 21. If you find the instructions extremely confusing, an alternative is to
use a package management system such as MacPorts. MacPorts and how to use it to install software are described on
Page 415.
At the end of this assgnment, you should have installed the following software:
1. TextWranger (jEdit for Windows/Ubuntu)
2. MySQL
3. Enthought Python Distribution
4. ImageMagick (compiling from source or using a package management system such as MacPorts)
#2 Creating your Duke home page
Requesting for AFS space All DUMC personnel with a NetID are eligible for AFS space (5GB) for hosting personal
web pages. However it is not available by default. Please email [email protected] to request for AFS space if
necessary to complete this assignment. It should be available to you within 24 hours of the request.
1. Create a filed called index.html in your text editor (TextWranger or jEdit) and type or copy the following text:
<html>
<head>
<title>My home page for PCfB</title>
</head>
<body>
Congratulations, you have successfully created your home page!
</body>
</html>

2. Use your NetID and password to log into WebFiles. Youll be connected to your home directory.
3. Click the Shared Spaces tab.
4. Under Your Personal Web Space, click Create public_html
5. Under Your Personal Web Space, click Upload to public_html and upload the index.html file you downloaded to
your desktop in Step 1.
6. To view your Web site, visit http://www.duke.edu/~NetID. (Replace NetID with your NetID but kep the ~)

4.2. Assignments

Practical Computing for Biologists, Release 1.0

4.3 References
The website for the textbook Practical Computing for Biologists.

4.3.1 The course website as a PDF

Workshop tutorials

4.3.2 NCBI eSearch

ESearch parameters

4.3.3 Unix
Unix Cheat Sheet

4.3.4 Regular expressions

Regular Expression Cheat Sheet

4.3.5 Python
Online tutorials
Learn Python the Hard Way: If you have found the learning curve for our exercises to be too steep, try the 52 exercises
at this site, which provide a much more gentle ramp. The author shares our philosophy that the only way to effectively
learn programming is by working on programming exercises. Dont be put off by the title - the exercises are not as
hard as the ones in the workshop - by the hard way the author just means learning by doing instead of learning by
reading.
Think Python - How to Think Like a Computer Scientist: Once you are comfortable with the basic syntax of Python
(e.g. from the book above), this book introduces you gently to the conceptual ideas you willl need to program effectively..
PyPi - A repository of software for the Python programming language
pypi
Useful packages for scientific computing
Python
Numpy
Scipy
Matplotlib
Sphinx

10

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

4.3.6 Relational Databases

SQLite
SQLite tutorial

4.3.7 Inkscape Tutorials

How to draw flow charts in Inkscape

4.3.8 Software used

TextWrangler
EPD Academic Version
MySQL
ImageMagick
ImageJ
1. Spellman P T, Sherlock, G, Zhang, M Q, Iyer, V R, Anders, K, Eisen, M B, Brown, P O, Botstein, D, Futcher, B.
Comprehensive identification of cell cycle-regulated genes of the yeast Saccharomyces cerevisiae by microarray
hybridization. Molecular biology of the cell, Vol. 9 (12): 3273-97, 1998. PubMed.
2. Duda, R O, Hart, P E & Stork, D G, Pattern Classification, John Wiley & Sons, Inc., 2001.
3. Cock, P J A and Antao, Tiago and Chang, J T and Chapman, B A and Cox, C J and Dalke, A and Friedberg, I
and Hamelryck, T and Kauff, F and Wilczynski, B and de Hoon, M J L, Biopython: freely available Python tools
for computational molecular biology and bioinformatics, Bioinformatics, Jun, 2009, 25, 11, 1422-3. PubMed.

4.3.9 Archival material

Snapshot of web content for workshop on June 1, 201

4.4 Participants
4.4.1 Registered
1. Will Williams <[email protected]>
2. Jessica Peel <[email protected]>
3. John Yi <[email protected]>
4. Alex Price <[email protected]>
5. Christopher J. Pierick <[email protected]>
6. Sandeep Dave <[email protected]>
7. Janet Staats <[email protected]>
8. Joe Saelens <[email protected]>
9. Anna Maria Masci <[email protected]>

4.4. Participants

11

Practical Computing for Biologists, Release 1.0

10. Herman Staats <[email protected]>

11. Adam Whisnant <[email protected]>
12. Pinghuang Liu <[email protected]>

4.4.2 Wait list

1. Luigi Racioppi <[email protected]> (non-CFAR)
2. Kelly Seaton <[email protected]>
3. Derrick Pulliam <[email protected]>
4. Guido Ferrari <[email protected]>

4.5 Installation and introduction

4.5.1 Introduction to PCfB
This workshop is intended to provide an introduction to the most useful tools for computation in biology. This includes
a basic command of the Unix shell, using text editors, regular expressions, scripting in the Python programming
language, data management/using a relational database, and creating vector graphics for scientific communication. As
there is a lot of new material to cover, we have created extensive documentation for each topic that will be accessible
at this website for your reference. Please let us know if any of the documentation is unclear or has errors - we want this
to be a useful resource for the future, and will fix documentation issues during the workshop itself as far as possible.
These are the specific topics that we will cover over the next few days
1. Basic Unix commands
2. Using a text editor and regular expressions
3. Remote computing and web page generation
4. Python Basics I
5. Python Basics II
6. Python Modules
7. NumPy and Matplotlib
8. Biopython I
9. Biopython II
10. Data management and relational databases
11. Data analysis with Python
12. Vector graphics with Inkscape

4.5.2 Software installation

But before we start, we just need to check that everyone has installed the required tools:
# An operating system based on Unix (Mac or Linux) # A text editor that understands regular expressions
# The Enthought Python distribution # A relational database system # Inkscape for vector graphics # A
web account on the Duke server with public_html access
12

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

If any of you have had trouble installing software, we will spend some time helping you to troubleshoot.

4.5.3 Feedback
Preparing for and running such a workshop takes a lot of time and effort. We are therefore very interested in any
feedback that you can provide that will help us improve. During the workshop, if you have any suggestions for
improvement, please let us know on the spot. Since this is a small class, we want the sessions to be highly informal
and welcome questions and interruptions.
We will probably run this course again in the future if you found it useful. It is also possible that we will run other
similar workshops, depending on interest. As a simple survey, what computational topics would you be interested in?
1. Practical programming for biologists - An intermediate course on the use of Python for scientific computation.
2. Practical statistics for biologists - An introduction to basic statistics in Python and R.
3. Practical data management for biologists - An introduction to creating and using relational database systems to
manage laboratory data.
4. Practical data visualization for biologists - An introduction to statistical and scientific graphics for exploratory
data analysis and scientific communication.
5. Modeling and simulations in biology - How to construct and simulate computational models of biological phenomena.
6. Others (please specify)

4.5.4 Pre-test
1. Do you know how to open a Unix shell/console/terminal on your computer?
2. How do you create a directory foo that has a subdirectory bar that has a subdirectory baz with a single
command?
3. How do you write a regular expression to find sequences that lie between specific restriction enzyme motifs?
4. What is the difference between ssh and scp?
5. How do you write a function in Python to plot a histogram of some data?
6. How do you use BioPython to get information from the NCBI databases?
7. What does this mean select f.name, b.value from foo f, bar b where f.foo_id =
b.foo_id;?
8. How can you estimate the 95% confidence intervals for a statistic without using any formulas?
9. Can you illustrate a conceptual biological model using a vector graphics program?
10. Can you write a program to summarize data from a typical laboratory spreadsheet?
Record your score from 0 to 10. We are curious to see if there is any improvement in your score by the end of the
workshop!

4.6 Basic Unix Commands

4.6.1 Working with the file system
Overview:
4.6. Basic Unix Commands

13

Practical Computing for Biologists, Release 1.0

cd - change directories
pwd - print working directory
mkdir - make directory
rmdir - remove directory
ls - list directory
cp - copy files
mv - move files
rm - remove files
Changing and Making Directories
pwd is a command that prints the current directory. Depending on how your shell is configured, your current directory or part of it is displayed in your prompt (the prompt is the bit in your shell that looks like this iMac:pcfb
cliburn$). Typically your shell starts you in your home directory, where you would have permissions to write and
create files. To change directories you would use cd, the change directory command.
[jacob@moku
/home/jacob
[jacob@moku
[jacob@moku
/tmp
[jacob@moku
[jacob@moku
/home/jacob

~]$ pwd
~]$ cd /tmp
/tmp]$ pwd
/tmp]$ cd
~]$ pwd

You can use cd without specifying a directory - this returns you to your home directory. You can also use ~ as an alias
for your home directory too. Creating directories uses the mkdir command. If you dont specify a full path (a path
starting with a /) it tries to create one in the current directory.
[jacob@moku ~]$ pwd
/home/jacob
[jacob@moku ~]$ mkdir foo
[jacob@moku ~]$ cd foo
[jacob@moku ~/foo]$ pwd
/home/jacob/foo
[jacob@moku ~/foo]$ mkdir /tmp/bar
[jacob@moku ~/foo]$ cd /tmp/bar
[jacob@moku /tmp/bar]$ pwd
/tmp/bar

If you need to make a deep hierarchy of directories all at once, you can use the -p argument to mkdir to create all the
necessary preceding directories.
[jacob@moku ~]$ pwd
/home/jacob
[jacob@moku ~]$ cd foo
foo: No such file or directory.
[jacob@moku ~]$ mkdir -p foo/bar
[jacob@moku ~]$ cd foo
[jacob@moku ~/foo]$ cd bar
[jacob@moku ~/foo/bar]$ pwd
/home/jacob/foo/bar

14

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

The rmdir command removes directories. Directories must be empty to be removed. Just like mkdir, if a full path
is not specified it tries to remove the directory from the current directory. Similar to mkdir, the -p argument tries to
remove all the preceding directories
[jacob@moku
[jacob@moku
[jacob@moku
[jacob@moku
[jacob@moku
[jacob@moku
rmdir: bar:

/tmp/bar]$ cd
~]$ rmdir /tmp/bar
~]$ rmdir -p foo/bar
~]$ mkdir bar
~]$ touch bar/foo
~]$ rmdir bar
Directory not empty

touch is a command that creates an empty file. We will find out about it when we look at working with files
Examining directories
Now that we understand directories, wed want to look at what files the directories contain. ls will list the files in a
directory.
[jacob@moku ~]$ ls
A.txt
B.txt
C.txt
[jacob@moku ~]$ ls bar
foo

bar

Just like mkdir, ls has several useful command line options. ls -l will list out all the extra properties of the directory
listed (file permissions, owner, last time modified). ls -a will list hidden files (those files whose name begin with a .).
ls -F will append directories names / and (along with other symbols after other special file types).
[jacob@moku ~]$ ls -l
total 5
-rw-r--r-- 1 jacob jacob 32 May 25 08:48 A.txt
-rw-r--r-- 1 jacob jacob 32 May 25 08:49 B.txt
-rw-r--r-- 1 jacob jacob 64 May 25 08:53 C.txt
drwxr-xr-x 2 jacob jacob
3 May 23 15:54 bar
[jacob@moku ~]$ ls -a
.
.cshrc
.mail_aliases
.rhosts
A.txt
bar
..
.login
.mailrc
.shrc
B.txt
.bash_history
.login_conf .profile
.ssh
C.txt
[jacob@moku ~]$ ls -laF
total 20
drwxr-xr-x 4 jacob jacob
16 May 27 12:11 ./
drwxr-xr-x 4 root
wheel
5 May 23 15:12 ../
-rw------- 1 jacob jacob
459 May 25 09:32 .bash_history
-rw-r--r-- 1 jacob jacob 1014 May 23 15:12 .cshrc
-rw-r--r-- 1 jacob jacob
257 May 23 15:12 .login
-rw-r--r-- 1 jacob jacob
167 May 23 15:12 .login_conf
-rw------- 1 jacob jacob
379 May 23 15:12 .mail_aliases
-rw-r--r-- 1 jacob jacob
339 May 23 15:12 .mailrc
-rw-r--r-- 1 jacob jacob
753 May 23 15:12 .profile
-rw------- 1 jacob jacob
284 May 23 15:12 .rhosts
-rw-r--r-- 1 jacob jacob
978 May 23 15:12 .shrc
drwx------ 2 jacob jacob
3 May 23 16:15 .ssh/
-rw-r--r-- 1 jacob jacob
32 May 25 08:48 A.txt
-rw-r--r-- 1 jacob jacob
32 May 25 08:49 B.txt
-rw-r--r-- 1 jacob jacob
64 May 25 08:53 C.txt
drwxr-xr-x 2 jacob jacob
3 May 23 15:54 bar/

4.6. Basic Unix Commands

15

Practical Computing for Biologists, Release 1.0

Working with files

Coping files uses the cp command, copying from the first argument (source) to the last (destination):
[jacob@moku ~]$ cp A.txt bar/A.txt

if the destination is a directory it copes the file into the directory.

[jacob@moku ~]$ cp A.txt bar/
[jacob@moku ~]$ ls bar
A.txt
foo

You can also copy multiple files at once. cp will copy all the files listed on the command line into the directory
specified in the last argument.
[jacob@moku ~]$ cp A.txt B.txt C.txt bar/
[jacob@moku ~]$ ls bar
A.txt
B.txt
C.txt
foo

Globbing will allow us to use many files at once rather than typing them all out explicitly. Globbing is a form of
wildcards.
Glob
*
?
[abc]

Effect
any number of any character
any single character
one of a, b, or c

[jacob@moku ~]$ cp *.txt bar/

[jacob@moku ~]$ ls bar
A.txt
B.txt
C.txt
foo

or
[jacob@moku ~]$ cp ?.txt bar/
[jacob@moku ~]$ ls bar
A.txt
B.txt
C.txt
foo

or even
[jacob@moku ~]$ cp [ABC].txt bar/
[jacob@moku ~]$ ls bar
A.txt
B.txt
C.txt
foo

with the -r command line argument you can recursively copy whole directories
[jacob@moku ~]$ cp -r bar foo
[jacob@moku ~]$ ls foo
A.txt
B.txt
C.txt
foo

Similar to the copy command is the move command mv.

[jacob@moku
[jacob@moku
[jacob@moku
[jacob@moku

~]$
~]$
~]$
~]$

mv
mv
mv
mv

A.txt foo
[BC].txt foo
foo/*.txt bar/
foo baz

To remove files, use the rm command. A word of caution, there is no trash can or waste basket. Removed files are
gone. It is very easy to accidentally shoot your self in the foot when blindly removing files.
[jacob@moku ~]$ rm bar/A.txt
[jacob@moku ~]$ rm bar/[BC].txt

16

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

[jacob@moku ~]$ ls bar

foo
[jacob@moku ~]$ rm -rf bar/

the -r command line flag removes files recursively, while -f attempts to ignore permissions on the file. The combination
of -r and -f flags can be useful to remove whole directory tree. BE VERY CAREFUL using -r and -f flags.

4.6.2 Working with file contents

cat - concatenate command
pipes and redirects - Why concatenate prints to the screen
globbing - Working with wildcards
less - a more sensible way to look at the contents of file
grep - searching for patterns in files
basics of regular expressions
Examining files
The cat command will display files on the screen
[jacob@moku ~]$ cat A.txt
This is file A.
It has 2 lines.
[jacob@moku ~]$ cat B.txt
This is file B.
It has
3 lines.

cat will also concatenate files to print to the screen.

[jacob@moku ~]$ cat A.txt B.txt
This is file A.
It has 2 lines.
This is file B.
It has
3 lines.

Using redirects allows us to save the concatenated file.

[jacob@moku ~]$ cat A.txt B.txt > C.txt
[jacob@moku ~]$ cat C.txt
This is file A.
It has 2 lines.
This is file B.
It has
3 lines.

> is a redirect to create a new file (and delete the old file if it exists). >> is the append redirect, while | (pipe) allow
you to send the output of one command as input to a new command.
[jacob@moku ~]$ cat [AB].txt
This is file A.
It has 2 lines.
This is file B.

4.6. Basic Unix Commands

17

Practical Computing for Biologists, Release 1.0

It has
3 lines.

While cat is useful for displaying small files, longer files would page off the screen quickly. To display longer files, a
page aware program will be used, less.
[jacob@moku ~]$ less <file name>

Common useful less keys

key
G
1G
#G
/foo
?foo
q

effect
Go to the last line
Go to the first line
Go to line number #
Search forward for foo
Search Backward for foo
Quit less

Quitting (or how to escape when you are lost)

You may at some point find you self lost, and your prompt doing interesting things you dont expect. Here are some
keys to try and get your prompt back in the state you expect.
q
Esc
ctrl-d (sends an end of file saying there is no more input)
typing Quit
typing exit
ctrl-c (sends a break, telling the program to abruptly halt)
Regexp, grep, and searching in files
The grep command allows you to tap into the powerful regular expression language to search the contests of file for
complex patterns.
[jacob@moku ~]$ grep poor Yorick hamlet.txt
Ham. Let me see. [Takes the skull.] Alas, poor Yorick! I knew him,

The first argument passed to grep is the pattern to search for, in the above example poor Yorick.
Regular expressions provide the ability to search beyond known text, using wildcards to build complex patterns.
key
.
+
*
^
$
[abc]

Meaning
any single character
one or more of the preceding character
zero or more of the preceding character
matches the beginning of the line
matches the end of the line
matches a singular character of a, b, or c

so to find all the lines beginning with the word HAMLET and end withs DEMARK

18

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

[jacob@moku ~]$ grep ^HAMLET.*DENMARK$ hamlet.txt

HAMLET, PRINCE OF DENMARK

Ever wonder how many lines in hamlet contain eight ls in them?

[jacob@moku ~]$ grep l.*l.*l.*l.*l.*l.*l.*l hamlet.txt
Till then sit still, my soul. Foul deeds will rise,
all welcome. Well een tot like French falconers, fly at
married already- all but one- shall live; the rest shall keep as
Clown. I like thy wit well, in good faith. The gallows does well.

4.6.3 Man Pages

Unix documentation is typically stored in man pages acces by the man command. Try typing man cat into the console
to see the manual page for the cat command. Note, man pages are notoriously terse, technical, and often confusing to
new users, so while learning it may be better to ask google instead, but if all you want is to know optional command
arguments the man page is the first place to look.

4.6.4 Exercise
Make a directory in your home folder named spam containing subfolders eggs, bacon, foo and bar and then remove
spam/foo and spam/bar

4.7 Using a text editor and regular expressions

4.7.1 What is a text editor?
Unlike a word processor, a text editor only handles plain text (i.e. no graphics or fancy formating). However, in return,
text editors provide powerful tools for manipulating text, including the ability to use regular expressions, highlight
differences between two or more files, and search and replace functionality on steroids. Since most data is available in
or can be exported to plain text, and computer programs are written in plian text, a text editor is one of the most basic
and useful applcations for anyone dealing with massive amounts of data. Text is universal - unlike binary formats (try
to open a WordPerfect 4.2 document), text documents are editable on any platform, and will still be readable in 30
years time.

4.7.2 First look

This session will asume that you are using the TextWrangler editor on a Mac. The functionality of most text editors is
very similar, and you should be able to follow along if you are using a different editor. We will take the lazy option
of familiarizing you to TextWrangler by using the official tutorial. Open TextWrangler and take a few minutes to
familiarize yourself with its anatomy - look at the menus, toolbars, icons etc. Most of this should be pretty familiar
to you from other programs. Open the file Lorem ipsum.txt in the Tutorial Examples/Lesson 3 folder. Now read the
description of the toolbar on Page 10 of the tutorial.

4.7.3 Mini-exercise
When you first open the Lorem ipsum.txt file, it looks like

4.7. Using a text editor and regular expressions

19

Practical Computing for Biologists, Release 1.0

Use the TextWrangler toolbar Text Options to get here

Then use the splitbar to split into two windows

Now drag the splitbar all the way to the top to get back a single window.

4.7.4 Finding and replacing

You can read all about TextWrangler in the tutorial at your leisure at home, but most things such as cut, copy and paste
and undo/redo work as expected. Since we have a lot to cover, well just skip ahead to lesson 7. Open the file Find
and Replace Sample.txt in the Lesson 7 folder and work through the exercise on page 39.
Next, open the Email Table.txt file in the Lesson 7 folder and work through the exercise on page 42 to learn how to do
search and replace on invisible characters.

20

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

4.7.5 Regular expressions

We now come to the main part of this session, which is how to use regular expressions (or regex) for text manipulation
described in pages 44 to 51 of the TextWrangler tutorial. You have already come across regular expressions in the
Unix session - now you will see how to use them to manipulate text files.
Well continue with the cats and dogs file just to get comfortable with the basic elements of regular expressions. Open
the Find dialog and check the Grep and Wrap around boxes. The Grep option tells TextWrangler that we are
using regular expressions, and wrap around that we want to do search and replace on the whole document regardless
of where our cursor/insertion point is. If at any point you get confused over the next few paragraphs, look at the
examples from page 47-49 for concrete illustrations of how to use regular expressions.
On page 45, there is a table of Wildcards. Put a period . in the Find box and click Next to see what matches. Keep
clicking Next - you will see that the . matches every character as advertised. Next try \s in the Find box and click
Next. What does it match? Repeat for all the Wildcards to get a good intuition of what each wildcard matches.
Now do the same with the Class patterns. Feel free to edit the text to include new words or numbers if you are curious
as to how they will be matched. Also experiment with creating your own class search patterns - for example, what
would a class to match DNA nucleotide symbols look like?
Now try the Repetition patterns *, *?, +, +? and ?. There is another repetition pattern {n, m} that means match at
least n but not more than m of the previous character or pattern. If n is left out, we have {,m} which means match no
more than m characters or patterns. If m is left out, we have {n,} which means match at least n occurrences. What is
the difference between * and *? or + and +?? When would you use the greedy and non-greedy versions?
Now figure out what the ^ and $ positional assertions and the alternation patterns mean. You can use parenthesis to
enclose alternations if you need to also match stuff before and/or after the alternation.

4.7.6 Mini-exercise
1. Construct a regular expression to find two or more consecutive vowels in the cats and dogs file.
2. Use a regular expression pattern to delete all punctuation from the example.
3. Use a positional assertion with the alternation pattern to find all cat, cats, dog or dogs that occur at the end a
sentence. You should find that the ends of sentences 1, 3, 5, 9 and 11 match.

4.7.7 Subpatterns and replace patterns

Regular expressions really begin to show their power when we use subpatterns and replace patterns. This can be a
little tricky to wrap your head around when encountered for the first time, so we start with some simple contrived
examples to give you some practice.
Subpatterns are simply any regular expression enclosed by parenthesis. For example, (.) is a subpattern that matches
any character, just like . by itself. However, we can refer to captured subpatterns to refer to whatever is in the
parenthesis. For example, what does the regex (.)\1 find? See if you can guess, then use TextWrangler to check if
you were correct.
We can also use the captured subpatterns in the replace pattern. Here again, \1 refers to the first regex in parenthesis,
2 to the second one etc, and & to refer to the full regular expression matched (not just an individual subpattern). As
an example, what does putting (.)\1([aeiou]+) in the Find box and &\2\1& do? Test it out in TextWrangler to
find out what word is changed and what it is changed to.

4.7.8 Exercise
Open the file Ch3observations.txt in the examples folder. It looks like this
4.7. Using a text editor and regular expressions

21

Practical Computing for Biologists, Release 1.0

13 January, 1752 at 13:53 -1.414 5.781 Found in tide pools

17 March, 1961 at 03:46 14 3.6 Thirty specimens observed
1 Oct., 2002 at 18:22 36.51 -3.4221 Genome sequenced to confirm
20 July, 1863 at 12:02 1.74 133 Article in Harpers
Write a regular expression to convert it into this:
1752 Jan.

13 13 53 -1.414 5.781

1961 Mar.

17 03 46 14 3.6

2002 Oct.

1 18 22 36.51 -3.4221

1863 Jul.

20 12 02 1.74 133

Hint: Construct a regular expression to match one single line. Look at the patterns that you must capture from
the original to perform the conversion. Construct the appropriate subpatterns to do so. Now construct the regular
patterns between the subpatterns to match the unwanted separating characters. When you have a regular expression
that matches a single line, check by clicking Next - the highlighted match should jump from one complete line to the
other. Now use the references \1, \2 etc, re-ordering if necessary, and adding in filler such as extra punctuation or
tabs to construct the desired replacement string. Save the regular expression by clicking the little g button on the Find
dialog box and clicking Save .... Now hit Replace and see if it does what you expect. If it works, hit replace a few
more times or click Replace All. If it doesnt work, Undo and try again.
If you are totally lost and about to pull all your hair out, the construction of the solution is described in detail on pages
38-40 of the PCfB textbook. However, you should not peek at the answer without trying for at least 15 minutes.

4.8 Remote computing and web page generation

Sometimes it is useful to be able to synchronize files or execute programs on a remote computer, or to transfer files
from one computer to another. This can be easily done from the command line very easily using the ssh, scp and
rsync commands.
The ssh (secure shell) command allows you to securely connect to a remote computer for which you have access. For
example, you can access your account on the Duke system:
eris:pcfb cliburn$ ssh [email protected]
[email protected] password:
Last login: Sat May 26 10:14:19 2012 from 152.3.189.147
##########################################################################
#
#
#
ATTENTION
ATTENTION
#
*****
*****
*****
#
#
# This is a Duke University computer system. This computer system,
#
# including all related equipment, networks and network devices
#
# (includes internet access) are provided only for authorized Duke
#
# University use. Duke University computer systems may be monitored for #
# all lawful purposes, including to ensure that their use is authorized, #
# for management of the system, to facilitate protection against
#
# unauthorized access, and to verify security procedures, survivability, #
# and operational security. During monitoring, information may be
#
# examined, recorded , copied , and used for authorized purposes. All
#
# information, including personal information, placed on or sent over
#
# this system may be monitored. use of this Duke University computer
#
# system, authorized or unauthorized, constitutes consent to
#
# monitoring. Unauthorized use of this Duke University computer system
#
# may subject you to criminal prosecution. Evidence of unauthorized use #

22

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

# collected during monitoring may be used for administrative, criminal, #

# or other adverse action. Use of this system constitutes consent to
#
# monitoring for all lawful purposes.
#
##########################################################################

Self provisioned systems are now available for remote usage through OITs Virtual Computing Lab servi
Mon May 28 00:44:43 EDT 2012
[ccc14@login5 ~]$ ls
AFSDocs public_html Sites
[ccc14@login5 ~]$

If you know the IP address of a Linux or Mac computer that you have login rights to, you can usually connect to it via
ssh. For example, this is how we typically access departmental servers or computing workstations from home. If you
work with the Duke Beowulf cluster, you will also use ssh to connect and run your programs remotely.
If you can ssh to a computer, you can also copy files to or from the remote computer using scp (secure copy). Here
is an example:
[ccc14@login5 ~]$ cat > remote.txt
This is my remote file on lgoin.oit.duke.edu
[ccc14@login5 ~]$ exit
logout
Connection to login.oit.duke.edu closed.
eris:pcfb cliburn$ scp [email protected]:~/remote.txt .
[email protected] password:
remote.txt
100%
45
0.0KB/s
eris:pcfb cliburn$ cat remote.txt
This is my remote file on lgoin.oit.duke.edu

00:00

If you wish to synchronize entire directory trees between computers, it is more efficient to use rsync which performs
data compression, only tranfers files that are differnet, and allows resuming of interrupted transfers. For example,
rsync is a simple way to back up your files to another computer.
eris:tmp cliburn$ rsync -avz foo [email protected]:~/
[email protected] password:
building file list ... done
foo/
foo/foo.txt
foo/bar/
foo/bar/bar.txt
foo/bar/baz/
foo/bar/baz/baz.txt
sent 376 bytes received 104 bytes
total size is 75 speedup is 0.16

73.85 bytes/sec

The flags -avz are short for --archive, --verbose and --compress. The --archive flag preserves symbolic links and is perfect for remote backups. As usual, you can look at man rsync if you want to know the details
of how rsync works.

4.8.1 Web page construction with sphinx

Sphinx (http://sphinx.pocoo.org/) is a Python tool to create documentation, but it is also great for creating highly
structured webpages with minimal effort. The entire workshop website was created with Sphinx.
We start by asking Sphinx to generate the initial directory for us with the sphinx-quickstart command. The
program will ask some configuration questions - you can just accept the defaults or give any sensible answer for now
- the options can all be changed later if necessary.
4.8. Remote computing and web page generation

23

Practical Computing for Biologists, Release 1.0

eris:tmp cliburn$ sphinx-quickstart homepage

Welcome to the Sphinx 1.1.2 quickstart utility.
Please enter values for the following settings (just press Enter to
accept a default value, if one is given in brackets).
Selected root path: homepage
You have two options for placing the build directory for Sphinx output.
Either, you use a directory "_build" within the root path, or you separate
"source" and "build" directories within the root path.
> Separate source and build directories (y/N) [n]:
Inside the root directory, two more directories will be created; "_templates"
for custom HTML templates and "_static" for custom stylesheets and other static
files. You can enter another prefix (such as ".") to replace the underscore.
> Name prefix for templates and static dir [_]:
The project name will occur in several places in the built documentation.
> Project name: Demo home page
> Author name(s): Cliburn Chan
Sphinx has the notion of a "version" and a "release" for the
software. Each version can have multiple releases. For example, for
Python the version is something like 2.5 or 3.0, while the release is
something like 2.5.1 or 3.0a1. If you dont need this dual structure,
just set both to the same value.
> Project version: 0.0
> Project release [0.0]:
The file name suffix for source files. Commonly, this is either ".txt"
or ".rst". Only files with this suffix are considered documents.
> Source file suffix [.rst]:
One document is special in that it is considered the top node of the
"contents tree", that is, it is the root of the hierarchical structure
of the documents. Normally, this is "index", but if your "index"
document is a custom template, you can also set this to another filename.
> Name of your master document (without suffix) [index]:
Sphinx can also add configuration for epub output:
> Do you want to use the epub builder (y/N) [n]:
Please indicate if you want to use one of the following Sphinx extensions:
> autodoc: automatically insert docstrings from modules (y/N) [n]:
> doctest: automatically test code snippets in doctest blocks (y/N) [n]:
> intersphinx: link between Sphinx documentation of different projects (y/N) [n]:
> todo: write "todo" entries that can be shown or hidden on build (y/N) [n]:
> coverage: checks for documentation coverage (y/N) [n]:
> pngmath: include math, rendered as PNG images (y/N) [n]:
> mathjax: include math, rendered in the browser by MathJax (y/N) [n]:
> ifconfig: conditional inclusion of content based on config values (y/N) [n]:
> viewcode: include links to the source code of documented Python objects (y/N) [n]:
A Makefile and a Windows command file can be generated for you so that you
only have to run e.g. make html instead of invoking sphinx-build
directly.
> Create Makefile? (Y/n) [y]:

24

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

> Create Windows command file? (Y/n) [y]:

Creating
Creating
Creating
Creating

file
file
file
file

homepage/conf.py.
homepage/index.rst.
homepage/Makefile.
homepage/make.bat.

Finished: An initial directory structure has been created.

You should now populate your master file homepage/index.rst and create other documentation
source files. Use the Makefile to build the docs, like so:
make builder
where "builder" is one of the supported builders, e.g. html, latex or linkcheck.

The next thing to do is to cd homepage to enter the directory that was just created for us and edit the conf.py
file to setup a configuraiton that we like. The only change to be made for now is to change the html_theme from
defautl to agogo to match our workshop website theme. The themes that come with Sphinx can be viewed at
http://sphinx.pocoo.org/theming.html.
The first page to edit is the index.rst file. The rst extension is for ReStructuredText, a simple plain text
markup language that is much easier to work with than HTML. Look at the primer on ReStructuredText at
http://sphinx.pocoo.org/rest.html to see examples of how to use it. Open the index.rst file in your text editor:
.. Demo home page documentation master file, created by
sphinx-quickstart on Mon May 28 01:12:56 2012.
You can adapt this file completely to your liking, but it should at least
contain the root toctree directive.
Welcome to Demo home pages documentation!
==========================================
Contents:
.. toctree::
:maxdepth: 2

Indices and tables

==================
* :ref:genindex
* :ref:modindex
* :ref:search

Since this is to be a home page rather than documentation page, we can simplify the structure. Edit the file so that the
last part looks like this:
Contents:
.. toctree::
:maxdepth: 2
:hidden:
Home <self>
research
publications

We want to keep the table of contents hidden, and have set up a simple structure where the home page (index.html)
4.8. Remote computing and web page generation

25

Practical Computing for Biologists, Release 1.0

links to a research.html and a publications.html file. Just as the index.html file will be generated by thiis index.rst file,
the other two files are also generated by a research.rst and publications.rst file that we write using ReStructuredText.
The full contents of the 3 rst files are included verbatim for reference:

4.8.2 index.rst
Cliburns very boring home page
==========================================
Stuff I do
------------------Tongue ribeye pig, tenderloin turducken salami frankfurter strip
steak. T-bone turducken meatball flank, beef ribs brisket corned
beef tail. Ball tip tongue flank beef ribs, biltong tri-tip salami
chicken sausage leberkas chuck tail. Kielbasa shankle pork chop
sirloin, leberkas bresaola tail. Ham hamburger venison sausage
biltong, pork loin brisket pig sirloin pastrami short loin shank
chicken. Pig andouille leberkas beef short loin ribeye turkey ham
hock. Cow ham kielbasa, capicola short ribs brisket shoulder
pancetta t-bone pork belly tri-tip pork loin tenderloin.
Ground round pork belly pastrami pork chop, drumstick corned beef
t-bone tail bresaola filet mignon meatloaf. Boudin spare ribs ham
hock short loin. Prosciutto ham hock sausage, biltong leberkas
turkey hamburger pork meatball bresaola pork belly. Shankle tri-tip
frankfurter ribeye leberkas ham hock, tongue beef ribs speck venison
pork chop andouille chuck. Rump pastrami bresaola, strip steak short
loin andouille pork chop beef boudin capicola bacon shank prosciutto
beef ribs swine. Meatloaf leberkas pancetta beef.
More stuff I do
-------------------Enim do boudin officia labore tail. Pork exercitation short ribs
deserunt laboris, tenderloin drumstick in dolor tongue sunt ex. Ham
hock t-bone exercitation pork loin non mollit. Jowl boudin magna
adipisicing in dolore. Brisket quis shoulder nostrud tempor
ea. Aliquip officia consequat deserunt, dolore nostrud est tri-tip
ut pancetta speck shank excepteur. Sausage cillum ground round velit
rump, dolore laboris.
Commodo consectetur ut, officia proident eu cillum jowl aute flank
sausage ut beef ribs. Deserunt occaecat pariatur elit. Pork chop ut
tempor, enim aliqua laborum cillum eiusmod t-bone occaecat aute
laboris labore. Ham hock turkey beef nostrud excepteur
dolor. Consectetur meatball chicken deserunt exercitation, corned
beef beef in short ribs ut ea velit beef ribs. Enim andouille in,
dolore ut meatball ea ut tail proident short ribs leberkas ground
round filet mignon.
Andouille sirloin chicken tempor aute, cow salami commodo dolore
leberkas culpa in ea esse. Id ground round tongue velit. Ex elit
minim sirloin fatback laboris. Irure andouille shankle cupidatat,
nostrud bresaola id shank do jowl. Swine sirloin pork loin,
prosciutto bresaola rump cillum in exercitation capicola.

26

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

Contents:
.. toctree::
:maxdepth: 2
:hidden:
Home <self>
research
publications

4.8.3 research.rst
Cliburns boring research page
=================================================
Current research interests
--------------------------------------1. Bacon
2. Pork rind
3. Trotters
.. image:: bacon.jpg
:width: 60%
Past research interests
---------------------------------------1. LOL cats
.. image:: Lolcat.JPG

4.8.4 publications.rst
Not really Cliburns publications
================================
First 5 hits on Pubmed search for "Sphinx"
--------------------------------------------1. Quadrature RF Coil for In Vivo Brain MRI of a Macaque Monkey in
a Stereotaxic Head Frame. Roopnariane CA, Ryu YC, Tofighi MR,
Miller PA, Oh S, Wang J, Park BS, Ansel L, Lieu CA, Subramanian T,
Yang QX, Collins CM. Concepts Magn Reson Part B Magn Reson
Eng. 2012 Feb;41B(1):22-27. Epub 2012 Feb 18. PMID: 22611340
[PubMed]
2. The place of general practitioners in cancer care in
Champagne-Ardenne. Tardieu E, Thiry-Bour C, Devaux C, Ciocan D,
de Carvalho V, Grand M, Rousselot-Marche E, Jovenin N. Bull
Cancer. 2012 May 1;99(5):557-562. PMID: 22522646 [PubMed - as
supplied by publisher]

4.8. Remote computing and web page generation

27

Practical Computing for Biologists, Release 1.0

3. Spontaneous Endometriosis in a Mandrill (Mandrillus sphinx).

Nakamura S, Ochiai K, Ochi A, Ito M, Kamiya T, Yamamoto H. J Comp
Pathol. 2012 Apr 18. [Epub ahead of print] PMID: 22520805
[PubMed - as supplied by publisher]
4. [Reality of healthcare access for migrant children in Mayotte].
Baillot J, Luminet B, Drouot N, Corty JF. Bull Soc Pathol
Exot. 2012 May;105(2):123-9. Epub 2012 Mar 1. French. PMID:
22383116 [PubMed - in process]
5. Craniodental features in male Mandrillus may signal size and
fitness: an allometric approach. Klopp EB. Am J Phys
Anthropol. 2012 Apr;147(4):593-603. doi: 10.1002/ajpa.22017. Epub
2012 Feb 10. PMID: 22328467 [PubMed - indexed for MEDLINE]

4.8.5 HTML Generation

With these files written, we now ask Sphinx to generate the HTML
eris:homepage cliburn$ make html
sphinx-build -b html -d _build/doctrees
. _build/html
Making output directory...
Running Sphinx v1.1.2
loading pickled environment... not yet created
building [html]: targets for 3 source files that are out of date
updating environment: 3 added, 0 changed, 0 removed
reading sources... [100%] research
Not really Cliburns publications
================================
looking for now-outdated files... none found
pickling environment... done
checking consistency... done
preparing documents... done
writing output... [100%] research
writing additional files... genindex search
copying images... [100%] bacon.jpg
copying static files... done
dumping search index... done
dumping object inventory... done
build succeeded, 1 warning.
Build finished. The HTML pages are in _build/html.

And now the directory looks like this:

eris:homepage cliburn$ ls
Lolcat.JPG
_templates
Makefile
bacon.jpg
_build
conf.py
_static
index.rst

make.bat
publications.rst
research.rst

4.8.6 Copy generated HTML to public_html on Duke server

Now all we need to do is to rsync the _build/html folder to the Duke server:

28

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

eris:homepage cliburn$ rsync -avz _build/html/ [email protected]:~/public_html/homepage

[email protected] password:
building file list ... done
created directory /afs/acpub/users/c/c/ccc14/public_html/homepage
./
.buildinfo
genindex.html
index.html
objects.inv
publications.html
research.html
search.html
searchindex.js
_images/
_images/Lolcat.JPG
_images/bacon.jpg
_sources/
_sources/index.txt
_sources/publications.txt
_sources/research.txt
_static/
_static/agogo.css
_static/ajax-loader.gif
_static/basic.css
_static/bgfooter.png
_static/bgtop.png
_static/comment-bright.png
_static/comment-close.png
_static/comment.png
_static/doctools.js
_static/down-pressed.png
_static/down.png
_static/file.png
_static/jquery.js
_static/minus.png
_static/plus.png
_static/pygments.css
_static/searchtools.js
_static/underscore.js
_static/up-pressed.png
_static/up.png
_static/websupport.js
sent 164472 bytes received 792 bytes
total size is 286369 speedup is 1.73

22035.20 bytes/sec

Now, if we navigate to http://www.duke.edu/~ccc14/homepage/, we will see the homepage and the links on the sidebar
to publications and research work as well.

4.8. Remote computing and web page generation

29

Practical Computing for Biologists, Release 1.0

4.9 Python Basics I

4.9.1 Why program?
Biology is increasingly data-rich. And the amount of data is growing exponentially, as I am sure you are all only too
well aware. Wouldnt it be really nice if you had some unpaid slaves that would help you sort through your data, check
for problems, find the most interesting bits, and generate pretty pictures for you to include in your next manuscript?
Well, you are in luck, for the computer is a slave factory, and the slaves are called programs. And today, I have some
slaves to give away ...

4.9.2 What can my slave do?

In this workshop, we will focus on 3 types of jobs for our slaves that are essential whenever there is a lot of data:
1. Data munging
2. Data analysis
3. Data visualization
Data munging means taking that error-riddled, color-coded, highly redundant Excel spreadsheet that biologists love
to create, and cleaning, parsing and proofing it so that it is suitable for analysis. There is a related skill of how to
30

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

create persistent data stores that allow you to slice and dice well-structured data that will be briefly touched upon
in the session on :doc:Data management and relational databasesI</database>, but will require
another full workshop to cover in any depth.
Data analysis is largely about how to do statistics. We will show very simple examples of analysis in :doc:Data
analysis with Python</analysis>, but the proper cultivation of the statistical way of thinking probably
requires not just another workshop, but returning to graduate school.
Finally, there are two main reasons for data visualization - the first is for exploratory data analysis, since the human
brain is highly optimized to detect patterns in pictures; and the second is for communicating results, since every
biologist Ive ever met is only ever interested in the figures in a paper and never the raw data. Making pictures from
data for exploratory analysis and communication are covered in NumPy and Matplotlib</numerics> and
the creation of schematics to illustrate concepts in :doc:Vector graphics with Inkscape</inkscape>.
But first - in order to tell your slave how to do these jobs, you need to think like a programmer.

4.9.3 Thinking like a programmer

Computers are stupid. And very literal. So to be able to program, we need to tell the computer what to do in very
simple language without any ambiguity. Almost all programs we will deal with involve only 5 basic operations:
1.
2.
3.
4.
5.

Get input data

Store data in variables
Do some calculation or check logic
Repeat
Generate some output

For now, we simply define a variable as a name we give to data so that we can retrieve it later. The other terms should
be familiar to everyone.
For example, here is a simple example that checks if a word is a palindrome:
1
2
3
4
5
6
7

# get input and store in variable word

word = raw_input("Enter a word: ")
# check if word is the same when reversed
if word == word[::-1]:
print "%s is a palindrome!" % word
else:
print "%s is a regular old word" % word

Here is another program that makes use of these 5 concepts:

1

"""Count the number of vowels in a name."""

2
3
4
5
6
7
8
9
10
11
12

# input data and store in variable name

name = raw_input("Whats your name? ")
# store count of vowels in variable num_vowels
num_vowels = 0
# repeat updating of vowel count
for vowel in aeiou:
# simple calculation (count occurences of vowel in string)
num_vowels += name.count(vowel)
# output result
print "Hello %s, there are %d vowels in your name." % (name, num_vowels)

Since the 5 operations are about all that a computer can do, even big complex projects must boil down to smaller tasks
that mix and match these operations. Essentially, if you know these 5 operations, you know how to program. The rest
are details.

4.9. Python Basics I

31

Practical Computing for Biologists, Release 1.0

4.9.4 Introducing Python

We will learn to program in a language called Python, named after the British comedy skit Monty Python. Python
is possibly the simplest programming language to learn and has an amazing range of libraries for just about any
biomedical data processing need, making it an ideal first language for biologists to learn. When you are comfortable
with Python, a very useful second language to learn is R, another open source language specialized for statistical
analysis.
Python is an interpreted language, meaning that instructions are executed as soon as you complete a programming
statement. To see the Python interpreter in action, we will use the IPython interpreter, which you can start by opening
a Terminal window and typing ipython, after which you will see this welcome message:
eris:pcfb cliburn$ ipython
Enthought Python Distribution -- http://www.enthought.com
Python 2.6.6 |EPD 6.3-2 (32-bit)| (r266:84292, Sep 23 2010, 11:52:53)
Type "copyright", "credits" or "license" for more information.
IPython 0.10.1 -- An enhanced Interactive Python.
?
-> Introduction and overview of IPythons features.
%quickref -> Quick reference.
help
-> Pythons own help system.
object?
-> Details about object. ?object also works, ?? prints more.

There is also a vanilla python interpreter, but ipython provides so many nice features such as Unix shell integration, tab completion, history etc that I hardly ever use the python interpreter. Try typing ? in ipython to see what
it offers. To exit the information screen, type q to quit. There are several ways to get more information about a Python
language feature - for example, what does the python range function do? Type help(range) or help range
or range? or ?range.

4.9.5 Python as a calculator

To get comfortable with ipython, lets just use it as a calculator:
In [1]: 1+2*3**2
Out[1]: 19
In [2]: 3/2
Out[2]: 1
In [3]: 3/2.0
Out[3]: 1.5
In [4]: 13 % 4
Out[4]: 1
In [5]: import math
In [6]: math.pi
Out[6]: 3.141592653589793
In [7]: math.e
Out[7]: 2.718281828459045
In [8]: math.pi/4
Out[8]: 0.7853981633974483

32

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

In [9]: math.sin(math.pi/4)
Out[9]: 0.7071067811865475
In [10]: math.asin(math.sin(math.pi/4))
Out[10]: 0.7853981633974482
In [11]: math.sqrt(16)
Out[11]: 4.0
In [12]: 16**0.5
Out[12]: 4.0

Note the gotcha in the [2] calculation - when both numerator and denominator are integers, the division operator
returns an integer, which might not be what you want. Make either numerator or denominator a float (a number with
a decimal point) as in [3] to get the usual answer.

4.9.6 Types
As we have already seen in the previous example, there is a difference between 2 and 2.0. In particular, they differ in
type - 2 is an integer, while 2.0 is a float. Types are necessary so that Python can distinguish between different kinds
of things that may have different behaviors. Here are the most commonly used basic types in Python:
1. Integers are natural numbers, ..., -3, -2, -1, 0, 1, 2, 3 ...
2. Floats are decimal numbers e.g. 0.01, 1e-6, math.pi etc
3. Bools are the values True and False
4. Strings are anything within single quotes, double quotes, or triple quotes such as hello, "hello",
hello and """hello""".
While the first 3 types are atomic, strings are actually sequences of characters, and we can retrieve characters at specific
postions by indexing and slicing. An example of how we can slice and dice sequences is useful here:
In [1]: s = "My first string"
In [2]: s[0]
Out[2]: M
In [3]: s[1]
Out[3]: y
In [4]: s[-1]
Out[4]: g
In [5]: s[0:2]
Out[5]: My
In [6]: s[3:8]
Out[6]: first
In [7]: s[3:8:2]
Out[7]: frt
In [8]: s[::-1]
Out[8]: gnirts tsrif yM

Note that in Python, we count from zero, not one. Note also that a negative index means count backwards from the
end of the sequence.
4.9. Python Basics I

33

Practical Computing for Biologists, Release 1.0

Another type of sequence that is ubiquitous in Python programs is the list, consisting of a sequence of other types
delimited by square brackets [ and ]. Unlike strings which only contain characters, list elements can be anything,
including other lists. Another difference between strings and lists is that the elements in a list can be changed by
assigning new values to them. In geek-speak, lists are mutable and strings are immutable. You may also see tuples
which are items separated by commas, and typically delimited by ( and ). For the most part, we can just consider
tuples to be immutable lists. A neat trick we can do with tuples is unpacking, perhaps easier demonstrated than
explained:
In the first example above, we unpacked the length 3 list into the variables a, b, c in a single statement. In the
second example, we swapped the contents of a and b.
Time for more experimentation in ipython:
In [1]: alist = [1,2,3.14,foo,bar,[a,b,True]]
In [2]: alist[5]
Out[2]: [a, b, True]
In [3]: alist[5][2]
Out[3]: True
In [4]: alist[1] = 99
In [5]: alist
Out[5]: [1, 99, 3.14, foo, bar, [a, b, True]]
In [6]: atuple = (1,2,3.14,foo,bar,[a,b,True])
In [7]: atuple[5]
Out[7]: [a, b, True]
In [8]: atuple[5][2]
Out[8]: True
In [9]: atuple[1] = 99
--------------------------------------------------------------------------TypeError
Traceback (most recent call last)
/Volumes/HD3/hg/pcfb/<ipython-input-9-5a8168f444b1> in <module>()
----> 1 atuple[1] = 99
TypeError: tuple object does not support item assignment
In [10]: astring = "hi there"
In [11]: astring[2] = x
--------------------------------------------------------------------------TypeError
Traceback (most recent call last)
/Volumes/HD3/hg/pcfb/<ipython-input-11-71b2376dea24> in <module>()
----> 1 astring[2] = x
TypeError: str object does not support item assignment

Here the difference between mutable and immutable is clearly shown. We can grow lists in several ways - using an
insert, and append and list concatenation. We can remove items from a list by using pop, del or assigning
a slice to the empty list.
In [1]: blist = []
In [2]: blist.append(1)

34

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

In [3]: blist.append(99)
In [4]: blist = blist + [3,4,5]
In [5]: blist
Out[5]: [1, 99, 3, 4, 5]
In [6]: blist[2:2] = [a,b,c]
In [7]: blist
Out[7]: [1, 99, a, b, c, 3, 4, 5]
In [8]: blist[2:4] = []
In [9]: blist
Out[9]: [1, 99, c, 3, 4, 5]
In [10]: blist.pop()
Out[10]: 5
In [11]: blist
Out[11]: [1, 99, c, 3, 4]
In [12]: del blist[3]
In [13]: blist
Out[13]: [1, 99, c, 4]

The final basic type we will look at is the dictionary. A dictionary consists of (key, value) pairs, where
the key is an immutable type (e.g. a number, a string, a tuple) and the value is anything. We retrieve the value in a
dictionary by using the associated key. Dictionaries are delimited by { and }. For example, we can make a dictionary
of email addresses:
In [1]: emails = {}
In [2]: emails[cliburn] = [email protected]
In [3]: emails[jacob] = [email protected]
In [4]: emails[cliburn]
Out[4]: [email protected]
In [5]: emails.keys()
Out[5]: [jacob, cliburn]
In [6]: emails.values()
Out[6]: [[email protected], [email protected]]
In [7]: emails
Out[7]: {cliburn: [email protected], jacob: [email protected]}

We can also think of dictionaries as fancy lists that are not restricted to consecutive integers for indexing. Note that
we create dictionaries with curly braces {} but assign element to and retrieve elements from dictionaries with square
brackets [key]. If the key is not found in the dictionary, Python will raise a KeyError exception and abort. To avoid
that, we can either check for the key before retrieval, tell Python to ignore KeyErrors in a try-except statement,
or return a default value using the get method instead of [] to access the dictionary. Here are more examples of
dictionary creation and usage:

4.9. Python Basics I

35

Practical Computing for Biologists, Release 1.0

In [1]: zip([a,b,c,d], [1,2,3,4])

Out[1]: [(a, 1), (b, 2), (c, 3), (d, 4)]
In [2]: adict = dict(zip([a,b,c,d], [1,2,3,4]))
In [3]: adict
Out[3]: {a: 1, b: 2, c: 3, d: 4}
In [4]: adict[b]
Out[4]: 2
In [5]: adict.get(e, 0)
Out[5]: 0
In [6]: adict
Out[6]: {a: 1, b: 2, c: 3, d: 4}
In [7]: adict.setdefault(e, 0)
Out[7]: 0
In [8]: adict
Out[8]: {a: 1, b: 2, c: 3, d: 4, e: 0}

Dictionaries can also be constructed from a list of (key, value) pairs (or 2-tuples). The zip function takes the first
element from list 1 and the first element from list 2 to make a tuple, then does the same for the second element etc until
one or both lists are exhausted. It is used here to construct a list of pair from two matching lists of keys and values.
The get method returns the default (second) argument when the key given by its first argument is not found in the
dictionary. The setdefault method does the same thing, but additionally inserts the new key / default value into
the dictionary if not found. If we want to ingore missing keys but just retrieve values for valid keys, we can wrap the
dictionary access in a try-except statement:
In [1]: adict
Out[1]: {a: 1, b: 2, c: 3, d: 4, e: 0}
In [2]: for ch in "ajfljldjajfeljad":
...:
try:
...:
print adict[ch]
...:
except KeyError:
...:
pass
...:
...:
1
4
1
0
1
4

The pass keyword means do nothing. Without the try-except statement, the program would crash with a
KeyError the first time ch was not found in the dictionary keys a, b, c, d, e.
You might have noticed that we sometimes used a funny notation with a dot . between names, for example
list.append(1). This is because Python is an object-oriented language, and these basic types are also classes.
We wont discuss classes here except to note that we use the dot notation to access values (attributes) and functions
(methods) associated with the class. In ipython, hit the tab key after the dot to see what types are available.
In [1]: blist
Out[1]: [1, 99, c, 4]

36

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

In [2]: [b for b in dir(blist) if not b.startswith(_)]

Out[2]:
[append,
count,
extend,
index,
insert,
pop,
remove,
reverse,
sort]

The code [b for b in dir(blist) if not b.startswith(_)] is a list comprehension to show all the normal
methods of the list class, filtering out methods that look like __xxx__. The methods with __ prefixes and suffixes
are special internal methods that we wont use in this workshop. You can use help to find out what extend, index
etc do. The count, reverse and sort methods are quite simple:
In [1]: numlist = [3,1,4,1,5,1,6,9]
In [2]: numlist.sort()
In [3]: numlist
Out[3]: [1, 1, 1, 3, 4, 5, 6, 9]
In [4]: numlist.reverse()
In [5]: numlist
Out[5]: [9, 6, 5, 4, 3, 1, 1, 1]

Strings have an even longer list of methods:

In [1]: quote = "My philosophy, like color television, is all there in black and white"
In [2]: [m for m in dir(quote) if not m.startswith(_)]
Out[2]:
[capitalize,
center,
count,
decode,
encode,
endswith,
expandtabs,
find,
format,
index,
isalnum,
isalpha,
isdigit,
islower,
isspace,
istitle,
isupper,
join,
ljust,
lower,
lstrip,
partition,
replace,

4.9. Python Basics I

37

Practical Computing for Biologists, Release 1.0

rfind,
rindex,
rjust,
rpartition,
rsplit,
rstrip,
split,
splitlines,
startswith,
strip,
swapcase,
title,
translate,
upper,
zfill]
In [3]: quote.lower()
Out[3]: my philosophy, like color television, is all there in black and white
In [4]: quote.upper()
Out[4]: MY PHILOSOPHY, LIKE COLOR TELEVISION, IS ALL THERE IN BLACK AND WHITE
In [5]: quote.split()
Out[5]:
[My,
philosophy,,
like,
color,
television,,
is,
all,
there,
in,
black,
and,
white]
In [6]: quote = .join(quote)
In [7]: quote.split(,)
Out[7]: [My philosophy, like color television, is all there in black and white]

4.9.7 Operators
We have already seen some operations, such as + and * for addition and multiplication. In addition to the numeric
operators, there are also Boolean operators and, or and not, comparison operators <, <=, >, >=, ==, !=, is and
is not, and some other operators we will not discuss here (e.g. bitwise operators). Most of these operators are quite
self-evident, and if not, experimentation in the interpreter will clarify what they do:
In [1]: True and False
Out[1]: False
In [2]: True or False
Out[2]: True
In [3]: not True
Out[3]: False

38

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

In [4]: 3 == 3
Out[4]: True
In [5]: 3 == 4
Out[5]: False
In [6]: 3 != 4
Out[6]: True
In [7]: 3 > 4
Out[7]: False
In [8]: 4 > 3
Out[8]: True
In [9]: None is None
Out[9]: True
In [10]: 0 is None
Out[10]: False
In [11]: 0 is not None
Out[11]: True

4.9.8 Control flow and loops

We now begin to get to the heart of programming - asking the computer to do mind-numbingly boring work many,
many times. The way to perform repetitions is by looping, but before we go there, we will first learn about how
to control the program flow with the if-else family of statements. Luckily, the if-else statement family does
exactly what youd expect it to do:
In [1]: if math > football:
...:
print Nerds rule
...: else:
...:
print Jocks rule
...:
Nerds rule

The structure of the if-else statement has the form:

if (condition is true):
do A
else:
do B

Here are some more examples of the if-else family:

In [1]: grade = None
In [2]: score = 86
In [3]: if (score > 93):
...:
grade = A
...: elif (score > 85):
...:
grade = B
...: elif (score > 70):
...:
grade = C

4.9. Python Basics I

39

Practical Computing for Biologists, Release 1.0

...: else:
...:
grade = D
...:
In [4]: grade
Out[4]: B

As you can see, the if statement can be used by itself without an else part with the understanding that if the condition
is not true, then nothing is done. If you need to make decisions based on many conditions, the if-[elif]-else
form is useful, where the final optional else statement will be executed if none of the others above it are true. Note
that the last example depends on the ordering of the conditions, and works because the if-elif-else statement
works from top to bottom. See if you can figure out why the code below doesnt work as intended:
In [1]: score = 86
In [2]: grade = None
In [3]: if (score > 70):
...:
grade = C
...: elif (score > 85):
...:
grade = B
...: elif (score > 93):
...:
grade = A
...:
In [4]: grade
Out[4]: C

OK, back to looping. There are two main ways to loop in Python using the for and while statements. The for
statement goes through a sequence of items one at a time, typically performing some work on that item as it iterates
over it. The examples below should make clear what a for loop does:
In [1]: range(10, 15)
Out[1]: [10, 11, 12, 13, 14]
In [2]: for number in range(10, 15):
...:
print number
...:
10
11
12
13
14
In [3]: for char in abcde:
...:
print char
...:
a
b
c
d
e
In [4]: for name in [adam, eve]:
...:
print name
...:
adam
eve

40

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

Remember that strings, lists, tuples and dictionaries are all sequences, and hence iterable. So we can use the for loop
on any of these. It is getting rather tedious to use the word sequence, so from now on, I will use lists, or strings or
dictionaries, but you should remember that the looping constructs work on all of them. Another Python idiom that is
sometimes useful in looping is the use of enumerate to keep track of position (or index) while looping over a list.
Here is how it is used:
In [1]: for i, name in enumerate([cliburn, jacob, adam]):
...:
print i, name
...:
0 cliburn
1 jacob
2 adam

A common use of the for loop is to create a new list from an old one. For example, here is how you can create a list of
all the squares from 10 to 15.
In [1]: squares = []
In [2]: for i in range(10, 16):
...:
squares.append(i**2)
...:
In [3]: squares
Out[3]: [100, 121, 144, 169, 196, 225]

Notice that to get the numbers [10,11,12,13,14,15], we call range(10, 16) since Python indexing includes the start but excludes the end.
We can now combine if checks with loops to filter lists that we are constructing, only adding items to our list if they
meet certain conditions. Suppose we wanted to only collect the squares of the odd numbers and discard the even one
in the previous example:
In [1]: oddsquares = []
In [2]: for i in range(10, 16):
...:
if i%2==1:
...:
oddsquares.append(i**2)
...:
In [3]: oddsquares
Out[3]: [121, 169, 225]

This process of looping over a sequence and collecting the items in a list, filtering by some condition if necessary, is so
common that Python has a short cut way of doing it known as list comprehension. Here is the nicer list comprehension
version of the above two examples:
In [1]: squares = [i**2 for i in range(10, 16)]
In [2]: squares
Out[2]: [100, 121, 144, 169, 196, 225]
In [3]: oddsquares = [i**2 for i in range(10, 16) if i%2==1]
In [4]: oddsquares
Out[4]: [121, 169, 225]

We can nest loops within each other - for example, to generate labels for a 96-well plate, we can do this list comprehension:

4.9. Python Basics I

41

Practical Computing for Biologists, Release 1.0

In [1]: wells = [%s%02d % (r, c) for r in ABCDEF for c in range(1, 13)]

In [2]: wells[:15]
Out[2]:
[A01,
A02,
A03,
A04,
A05,
A06,
A07,
A08,
A09,
A10,
A11,
A12,
B01,
B02,
B03]

It might be clearer to understand what is happening using the longer version of creating an empty list, then using
nested for loops:
In [1]: wells = []
In [2]: for r in ABCDEF:
...:
for c in range(1,13):
...:
wells.append(%s%02d % (r, c))
...:
In [3]: wells[:15]
Out[3]:
[A01,
A02,
A03,
A04,
A05,
A06,
A07,
A08,
A09,
A10,
A11,
A12,
B01,
B02,
B03]

Dont worry about the funny %02d and %s bits in the code. We will explain string interpolation in the next
session. Congratulations! You have learnt the basic building blocks of a python program. Once you complete the
exercise below, we will finish off our Python crash course with I/O and creating your own functions.

4.9.9 Storing and re-using programs

Using ipython is great for learning because of the instant feedback, but at some point you will want to save your
code to use for another day. To do so, we use our text editor to write code in a file that ends with the extension .py.
Open up a new page in TextWrangler, enter the following code:

42

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

print "I am an expert programmer"

then save it as expert.py. Now open a new terminal, and run the program like so:
eris:~ cliburn$ python expert.py
I am an expert programmer

You can also run Python programs from within ipython with the run keyword:
In [1]: run expert.py
ERROR: File expert.py not found.

For the above two programs to run, you need to be in the same directory as expert.py. Later, Jacob will show you
how to run programs in arbitrary locations.

4.9.10 Exercise
It is claimed by an interviewer that the majority of computer science graduates cannot write a correct solution to
this problem (http://imranontech.com/2007/01/24/using-fizzbuzz-to-find-developers-who-grok-coding/). Are you better than the average computer science graduate?
Write a program that prints the numbers from 1 to 100. But for multiples
of three print Fizz instead of the number and for the multiples of five
print Buzz. For numbers which are multiples of both three and five print
FizzBuzz.
You should save your program in a text editor and execute it as shown above. You can experiment within
ipython, and copy and paste working code from ipython into the text editor. The instruction history -n
startline:stopline shows you code in your history without the line numbers that is convient for cutting and
pasting. Do it in stages - first, how do you print the numbers from 1 to 100? Next, how do you find multiples of 3?
How do you change what is printed if the number is a multiple of 3? And so on ...

4.10 Python Basics II

We left out the while looping construct yesterday. It is used like this:
initialize condition
while (condition is True):
do stuff

Typically, the condition is updated within the body of the while statement such that it eventually becomes false. A
simple example follows:
In [1]: i = 0
In [2]: while (i < 5):
...:
i = i+1
...:
print i
...:
1
2
3
4
5

4.10. Python Basics II

43

Practical Computing for Biologists, Release 1.0

4.10.1 Review and warm-up exercises

1. Open a terminal. What directory are you in? List the files in your current directory. Make the following new directories - folder1/folder2/folder3. Navidgate to folder2 and create a file catted.txt
using cat with the words I made this!. Delete folder1, folder2 and folder3.
1. Open your text editor. Load the file sequence1.txt. Find the sequences that would be
cut by the EcoRI restriction enzyme that recognizes sequneces flanked on both sides by
GAATTC.
2. Create a list of cubes for all the positive inteters less than 10 (i.e. [1,8,27,...,729]) using a)
a for loop and b) a list comprehension.
3. Find all the characters that are used exactly once in the sentence A person who never
made a mistake never tried anything new. Ignore case, so A and a would be counted
as 2 occurrences of a.
4. Find the second largest number in this list [ 9, 61, 2, 79, 58, 87, 68, 83, 61, 13]
5. Make a dictionary from these two lists, using the author as the key and the quote as the
value:
authors = ["Albert Einstein", "Richard Feynman", "Charles
Darwin"]
quotes = ["Any man who can drive safely while kissing a
pretty girl is simply not giving the kiss the attention it
deserves", "There is a computer disease that anybody who works
with computers knows about. Its a very serious disease and
it interferes completely with the work. The trouble with
computers is that you play with them!", "I have tried lately
to read Shakespeare, and found it so intolerably dull that it
nauseated me"]
1. Write a program that generates this list [0,1,2,3,4,5,5,4,3,2,1,0]. Your program should
only contain a single integer.
2. Write a program that uses while and raw_input and simply repeats the question who
wins? until you type the words you win. Here is a session with the program in the
terminal:
eris:examples cliburn$ python winner.py who wins? me who wins? cliburn who wins? someone else who wins? you win

4.10.2 String Interpolation

We start this session by learning how to construct nicely formatted strings with holes where we want to insert variables.
The most convenient way to do this in python is by string interpolation. String interpolation uses codes starting with
the % symbol as placeholders for inserting variables within the string. The codes that are most useful are:
1. %s for strings
2. %d for integers
3. %f for floats
So for example, we can have this string:
In [1]: "%s took %d courses last year and had a %f GPA" % (Tome, 4, 3.2)
Out[1]: Tome took 4 courses last year and had a 3.200000 GPA

44

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

Note that the variables to be inserted into the string are given as a tuple following the % separator. However, the default
formatting leaves something to be desired. For numbers, we can specify the minimum width, as well as the number of
decimal places when the number is a float.
In [1]: %4d % 123
Out[1]: 123
In [2]: %4d % 12345
Out[2]: 12345
In [3]: %5f % 3.14
Out[3]: 3.140000
In [4]: %5.2f % 3.14
Out[4]: 3.14

Sometimes it is also convenient to pad strings or change alignment so rows line up nicely using the flags 0 (left pad
with zeros), (left pad with space), - (left align) and + (add sign character).
In [1]: %05d % 23
Out[1]: 00023
In [2]: % 5d % 23
Out[2]:
23
In [3]: %-5d % 23
Out[3]: 23

In [4]: %+5d % 23
Out[4]: +23

Simple tables are often created using loops and string interpolation. For example, here is the code to print out the
layout of a 96 well plate:
In [1]:
...:
A01 A02
B01 B02
C01 C02
D01 D02
E01 E02
F01 F02

for r in ABCDEF: print .join([%s%02d % (r, c) for c in range(1, 13)])

A03
B03
C03
D03
E03
F03

A04
B04
C04
D04
E04
F04

A05
B05
C05
D05
E05
F05

A06
B06
C06
D06
E06
F06

A07
B07
C07
D07
E07
F07

A08
B08
C08
D08
E08
F08

A09
B09
C09
D09
E09
F09

A10
B10
C10
D10
E10
F10

A11
B11
C11
D11
E11
F11

A12
B12
C12
D12
E12
F12

The join method of a string joins together all the strings in a list, separated by the original string. In this case the
original string is a space , so all the strings in the list comprehension will be joined with spaces separating them
before being printed. Take your time to deconstruct this short example - it pulls together many concepts - looping, list
comprehension, string interpolation and the use of the string method join.

4.10.3 Mini-exercise
1. Write a program that produces these 2 lines of output from range(1,11):
0001 0002 0003 0004 0005 6.00 7.00 8.00 9.00 10.00
2. Write a program that starts with range(1, 6) and ends up with this string 1-one-thousand-2-one-thousand-3-onethousand-4-one-thousand-5, using a list comprehension, the str() function and a string join.

4.10. Python Basics II

45

Practical Computing for Biologists, Release 1.0

4.10.4 Reading from and writing to files

We open files using the built-in open function. We need to tell the function if the file is to be used for reading, writing
or appending with the r, w and a flags. When a file is opened for reading (the default), its contents cannot be altered.
When a file is open for writing, if there is an existing file with the same name, its contents are deleted and you can
write new content to the file. Opening for appending does not delete pre-existing file contents, but allows addition of
new content appended to the existing contents. If you work with files from different operating systems, an additional
useful flag is U for universal that handles differences between how Unix, Macs and Windows systems deal with line
endings. All the following do the same thing - open the file sequnce1.txt in the examples directory for reading.
The last version is the most robust, and will work regardless of whether sequnce1.txt was created on a Unix, Mac
or Windows system, while the others only work if the file was created on the same platform as you are currently using.
In [1]: fin = open(examples/sequence1.txt)
In [2]: fin = open(examples/sequence1.txt, r)
In [3]: fin = open(examples/sequence1.txt, rU)

Opening files for writing or appending is similar, but replace the r in the argument with w or a. Remember if you
open the file sequence1.txt with the w flag, the current contents are gone forever.
OK. Now we will open a file for writing, write some lines, close it, open again for appending more lines, close it, and
finally open again for reading.
In [1]: graffiti = \n.join([Roses are red, Violets are blue, The dog is pregnant, Thanks to
In [2]: fo = open(graffiti.txt, w)
In [3]: fo.write(graffiti)
In [4]: fo.close()

Here, we write some lines of doggerel in a list, join them as separate lines with the newline separator \n, then write it
to a file called directory.txt that has been opened for writing. Sometimes, you will see another newer idiom for
opening files:
In [1]: with open(graffiti.txt, w) as fo:
...:
fo.write(graffiti)
...:

The difference is that when using the with statement, you dont need to remember to close the file handler. The
operating system limits the numbers of file handlers that are available, and exceeding the number may lead to a system
crash. Closing the file frees up the resource, but it is easy to forget to do so in more complicated programs, hence the
availability of the with statement. Either way is fine. You can see whats in the file by using less graffiti.txt
either in ipython or on the command line.
Lets add another line for the author of the poem:
In [1]: fo = open(graffiti.txt, a)
In [2]: fo.write(\n + by anonymous college toilet poet)
In [3]: fo.close()

Note that we add a newline \n before the attribution string so that it appears on a separate line.

46

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

4.10.5 Mini-exercise
1. Find the AT/GC ratio in sequence1.txt.
2. Find all palindromes of length = 9 in sequence1.txt and save them to a file called palindromes.txt.
3. Now, re-open palindromes.txt and append all palindromes of length 8 to the file.

4.10.6 Reading files with read() and readlines()

We can also store the contents of the file read in by using the read and readlines methods for further processing
later. The difference is that read returns the content as a single string, while readlines returns it as a list of lines.
In [1]: poem = open(graffiti.txt, rU).read()

In [2]: poem
Out[2]: Roses are red\nViolets are blue\nThe dog is pregnant\nThanks to you\nby anonymous college to
In [3]: poem = open(graffiti.txt, rU).readlines()
In [4]: poem
Out[4]:
[Roses are red\n,
Violets are blue\n,
The dog is pregnant\n,
Thanks to you\n,
by anonymous college toilet poet]

We can now process the string in poem1 or the list in poem2 as necessary.

4.10.7 Exercise
1. Convert the contents of the file graffit.txt to all uppercase letters. That is, calling cat or less on graffit.txt
should look like this before and after your program is run:
BEFORE
eris:pcfb cliburn$ cat graffiti.txt
Roses are red
Violets are blue
The dog is pregnant
Thanks to you
by anonymous college toilet poet

AFTER
eris:pcfb cliburn$ cat graffit.txt
ROSES ARE RED
VIOLETS ARE BLUE
THE DOG IS PREGNANT
THANKS TO YOU
BY ANONYMOUS COLLEGE TOILET POET
eris:pcfb cliburn$

2. Count the number of times each word appears in hamlet.txt found in the
examples folder. For, we define a word to be any string of characters
that is separated by white space (space, tab, newline). We also ignore

4.10. Python Basics II

47

Practical Computing for Biologists, Release 1.0

case - so ABC is the same word as abc.

punctuation before doing the word count.

For extra credit, strip all

1. Open the file hamlet.txt and assign its contents to a variable as a single string
2. Convert the string to lower case
3. Remove all punctuation characters from the string (punctuation characters are !#$%&()*+,./:;<=>?@[\]^_{|}~, which you can also find in the string module)
4. Split the string into a list of words, where a word is defined to be any sequence of characters separated by white
space
5. Create an empty dictionary to store word counts
6. Loop over the list of words and increment the dictionary count for that word by 1
7. Print the number of occurrences of hamlet in Hamlet
8. Close the file if necessary

4.10.8 Writing your own functions

We are finally in the home stretch for the Python programming module. Learning to write your own functions will
greatly increase the complexity of the programs that you can write. A function is a black box - it takes some input,
does something with it, and spits out some output. Functions hide details away, allowing you to solve problems at a
higher level without getting bogged down. For example, consider the built-in sum function:
In [1]: numbers = [1,6,23,8,1,2,90]
In [2]: sum(numbers)
Out[2]: 131

The use of the built-in sum function hides the details of having to initialize the sum to zero and looping over each
number while adding that number to the sum variable. While Python comes with many useful built-in functions, sooner
or later, you will need to write your own functions. As you will see, writing your own functions is really simple. Lets
write our version of the sum function and a product function that when given a sequence of numbers, returns the
product rather than the sum of numbers. We will store save the functions in examples/functions.py.
def sum(xs):
"""Given a sequence of numbers, return the sum."""
s = 0
for x in xs:
s += x
return s
def prod(xs):
"""Given a sequence of numbers, return the product."""
s = 1
for x in xs:
s *= x
return s

A typical function looks like this:

def function_name(function_arguments):
"""Optional string describing the function."""
statements ...
return result

48

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

4.10.9 Mini-exercise
1. Write a function that returns the cumulative sum of numbers in a list. For example, if the function is given the
list [1,2,3,4,5], it should return the list [ 1, 3, 6, 10, 15].
2. Write a function fib that generates the first n Fibonacci numbers. The Fibonacci numbers are the sequence
[1,1,2,3,5,8,13,...], where each successive number is the sum of the two preceding numbers. Here are some
results that your function should give:
In [1]: fib(1)
Out[1]: [1]
In [2]: fib(2)
Out[2]: [1, 1]
In [3]: fib(3)
Out[3]: [1, 1, 2]
In [4]: fib(4)
Out[4]: [1, 1, 2, 3]
In [5]: fib(10)
Out[5]: [1, 1, 2, 3, 5, 8, 13, 21, 34, 55]

4.10.10 Importing a function from a file (module)

Now we can use the prod function in ipython or other programs just like a built-in function once we import the
functions module we have just written. If we are not in the same directory as functions.py, we also need to
tell the Python interpreter where to find it by adding its location to the search path sys.path (which is just a list
of directories that Python looks for modules). The way to call a function is to give the function name followed by
parenthesis with values for the number of arguments expected:
In [1]: import functions
In [2]: xs = [1,2,3,4]
In [3]: s, p = functions.sum(xs), functions.prod(xs)

If calling functions.prod is too verbose for you, you can modify the import statement like so:
Just be aware that this will make any existing function with the name prod inaccessible. So for instance, if we
used from functions import sum, we would no longer have access to the built in sum function. Whereas
if we used import functions, we can choose which function to use - sum will use the built-in function, while
functions.sum will use our function. We recommend using the full name all the time to avoid such name clashes,
using a shorter alias for the imported module with the as keyword if you are really lazy.
In [1]: import functions as f
In [2]: f.prod(xs)
Out[2]: 24

4.10.11 Functions are first class objects

In Python, functions can be treated like any other object - we can assign them to variables, use them as values in
dictionaries, use them as arguments to other functions etc.

4.10. Python Basics II

49

Practical Computing for Biologists, Release 1.0

In [1]: foo = functions.sum

In [2]: foo([1,2,3])
Out[2]: 6
In [3]: func_dict = {plus : functions.sum, times : functions.prod}
In [4]: xs = [1,2,3,4]
In [5]: func_dict[plus](xs)
Out[5]: 10
In [6]: func_dict[times](xs)
Out[6]: 24

There is another way to write short throwaway functions for one-time use that is much terser using lamba or anonymous functions:
In [1]: f = lambda x: x*x
In [2]: f(3)
Out[2]: 9

This use of lambda is typically seen in the context of the built-in higher order functions (functions that take functions
as arguments) map and filter.
In [1]: filter(lambda x: x % 2==0, range(10))
Out[1]: [0, 2, 4, 6, 8]
In [2]: map(lambda x: x**2, range(5))
Out[2]: [0, 1, 4, 9, 16]

In general, Python programmers prefer to use defined rather than anonymous functions, and the use of list comprehensions rather than map and filter as they are more explicit and easier to understand, but you may come across
lambda, map and filter in books or on the web.

4.10.12 Mini-Exercise
1. Replace the filter and map functionality in the above example using list comprehension.
2. Rewrite f = lambda x: x**2 as a regular function also called f using def.

4.10.13 Function arguments

We can define functions with more than one argument, as well as give default values to the arguments. In turn, when
calling a function, we can supply the arguments either by position or by name. Arguments with default values do not
need to be supplied when calling a function, but if provided, will overwrite the default values.
In [1]: def f(a, b, c=3, d=100):
...:
print a, b, c, d
...:
In [2]: f(1,2)
1 2 3 100
In [3]: f(1,2,3,4)

50

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

1 2 3 4
In [4]: f(d=1, c=2, b=3, a=4)
4 3 2 1

Warning: When you assign a list or a dictionary as a default value for an argument, the list is created at the same time
the function is declared, and persists over subsequent function calls if not overwritten. That is probably not what you
intended - if you do not want the default list to persist, you have set the default to None in the argument, then set it to
the empty list in the function after checking that it has not been assigned. An example should make this clear:
# we set b to have a default of an empty list
In [1]: def f(a, b=[]):
...:
b.append(a)
...:
print a, b
...:
# but the behavior is rather counter-intuitive
In [2]: f(2)
2 [2]
In [3]: f(3)
3 [2, 3]
# if we over-write the default argument, everything is OK
In [4]: f(3, [1,2])
3 [1, 2, 3]
# this is the way to get the non-persistent behavior
In [5]: def f(a, b=None):
...:
if b is None:
...:
b = []
...:
b.append(a)
...:
print a, b
...:
In [6]: f(2)
2 [2]
In [7]: f(3)
3 [3]

4.10.14 Exercise
1. Write a function that finds palindromic sequences of length k from a string, and use it to find all palindromic
sequences of length 9 in sequence1.txt in the examples folder. The function should take 2 arguments, the string
and k, the palindrome length
2. Write a program that plays the childrens guessing game with you. Running the program and playing with it
looks like this:
eris:examples cliburn$ python guessing.py
Im thinking of a number between 1 and 100. Guess what it is!
Guess a number: 50
Too small
Guess a number: 75
Too large
Guess a number: 63

4.10. Python Basics II

51

Practical Computing for Biologists, Release 1.0

Too large
Guess a number: 56
Too small
Guess a number: 60
Too large
Guess a number: 58
Youve guessed it! The number is 58

The first 5 lines of the program look like this:

import random
number = random.randint(1, 101)
guess = None
print "Im thinking of a number between 1 and 100. Guess what it is!"
while guess != number:
YOUR CODE HERE

4.11 Python Modules

Python comes with many libraries covering a large variety of functions. During this section we will look at some of
them.
Recall that libraries are accessed using the import statement. You can see what object a library contains by using
dir(<module>) in ipython, and looking at <module>.<object>.__doc__ string
In [1]: import math
In [2]: print math.exp.__doc__
exp(x)
Return e raised to the power of x.

4.11.1 os module
The os module provide a platform independent way to work with the operating system, make or remove files and
directories.
In [1]: import os
In [2]: print os.getcwd()
/home/jacob
In [3]: os.chdir(/home/jacob/baz)
In [4]: os.getcwd()
Out[4]: /home/jacob/baz
In [5]: os.remove(foo)
In [6]: os.chdir(/home/jacob)
In [7]: os.rmdir(baz)

52

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

4.11.2 csv module

The csv module provides readers and writers for comma separated value data.
In [1]: import csv
In [2]: f = open(monty_python.csv, w)
In [3]: cfw = csv.writer(f, dialect=excel)
In [4]: cfw.writerow([spam,spam,eggs,spam])
In [5]: cfw.writerow([spam,spam,spam and eggs, spam])
In [6]: cfw.writerow([toast])
In [7]: f.close()
In [8]: cfr = csv.reader(open(monty_python.csv, rU), dialect=excel)
In [9]: for
...:
...:
spam, spam,
spam, spam,
toast

row in cfr:
print , .join(row)
eggs, spam
spam and eggs, spam

If youd prefer a different separator than commas the delimiter optional argument can be used
In [1]: import csv
In [2]: f = open(tabbs.csv, rU)
In [3]: csvfile = csv.reader(f, delimiter=\t)
In [4]: for row in csvfile:
...:
print row
...:
[1, 2, 3]
[2, 3, 4]
[4, 5, 6]

often youll want to skip the first row in a csv file, and a simple way to do that is
import csv
f = open(test_scores.csv, rU)
csvfile = csv.reader(f)
header = False
for row in csvfile:
if not header:
header = True
else:
print row
f.close()

4.11. Python Modules

53

Practical Computing for Biologists, Release 1.0

csv excercise
use the test_scores.csv file to calculate the average score (colum 4) for each sex (column 2)

4.11.3 sys.argv
The sys module contains many objects and functions for dealing with how python was compiled or called when
executed. Most significantly is argv, which is a list containing all the parameters passed on the command line when
python executed, including the name of the python program in position 0. Note that all the elements of sys.argv are
strings - if you want a number, you will have to convert it using int() or float(). For example, if you want to
assign the argument at position 1 as an integer variable, you can use n = int(sys.argv[1]).
import sys
if len(sys.argv) > 1:
print sys.argv
else:
print no arguments passed
[jacob@moku ~]$ python argv_example.py foo bar baz
[argv_example.py, foo, bar, baz]

sys.argv exercise
write a program that takes two arguments, your first name, and your age, and then prints out your name and the year
you were born.

4.11.4 glob module

The glob module proves the glob function to perform file globbing similar to what the unix shell provides
In [1]: import glob
In [2]: for file in glob.glob(./*.txt):
...:
print file
...:
C.txt
B.txt
A.txt
hamlet.txt

4.11.5 math module

The math module provides common algebra and trigonometric function along with several math constants. Note
that the trigonometric functions work in radians rather than degrees. You can convert from radians to degrees with
math.degrees and vice versa with math.radians.
In [1]: import math
In [2]: math.e
Out[2]: 2.718281828459045
In [3]: math.pi
Out[3]: 3.141592653589793

54

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

In [4]: math.log10(100)
Out[4]: 2.0
In [5]: math.log(math.e)
Out[5]: 1.0
In [7]: math.cos(math.pi)
Out[7]: -1.0
In [10]: math.exp(1)
Out[10]: 2.7182818284590455
In [11]: math.pow(5,2)
Out[11]: 25.0
In [12]: math.sin(math.pi)
Out[12]: 1.2246467991473532e-16

for those wondering about line 12, see floats

4.11.6 re (regular expressions)

The re module provides access to powerful regular expressions. Use re.search to determine if a pattern exists in a
string, or use re.findall to search for all instances of a pattern in a string.
In [1]: import re
In [2]: m = re.search(games, all fun and games)
In [3]: m.group()
Out[3]: games
In [4]: re.findall([1234567890]+,12 drummers drumming, 11 pipers piping, 10 lords a-leaping)
Out[4]: [12, 11, 10]

4.11.7 datetime and time

The datetime module provides time and datetime objects, allowing easy comparison of times and dates.
In [1]: import datetime
In [2]: a = datetime.time(8,30)
In [3]: b = datetime.time(9,45)
In [4]: b>a

The time module provides simple estimates for how long a command takes.
In [1]: import time
In [2]: a = time.time()
In [3]: time.sleep(10)
In [4]: b = time.time()

4.11. Python Modules

55

Practical Computing for Biologists, Release 1.0

In [5]: print b-a

10.743445158

4.11.8 pickling and unpickling

The pickle module provides a way to save python objects to a file that you can unpickle later in a different program.
This allows the saving and loading of complex objects such as classes, dictionaries etc.
In [1]: import pickle
In [2]: adict = {a: [1,2,3], b : [4,5,6]}
# pickle.dump takes as arguments the object to be pickled and a file handler
In [3]: pickle.dump(adict, open(adict.pic, w))
# delete adict to see if loading from a pickled file works
In [4]: del adict
# adict has been deleted, so we get an error message
In [5]: adict
--------------------------------------------------------------------------NameError
Traceback (most recent call last)
/Users/cliburn/hg/pcfb/<ipython-input-5-5f470a13239f> in <module>()
----> 1 adict
NameError: name adict is not defined
# load adict from pickled file
In [6]: adict = pickle.load(open(adict.pic))
In [7]: adict
Out[7]: {a: [1, 2, 3], b: [4, 5, 6]}

4.11.9 pypi (formerly cheese shop)

The python community maintains a database of third party python packages. This database lives at pypi. Many of
these packages can be installed using pip or easy_install. In the example below, I first use easy_install to install pip,
then use pip to install xlrd (a module for reading XLS files). The -U flag means that the program should update the
installation if the module is already installed. Finally, we can also use pip to uninstall modules (but not easy_install,
which is why we prefer to use pip).
iMac:pcfb cliburn$ easy_install -U pip
Searching for pip
<SNIP>
Installed /Users/cliburn/Library/Python/2.7/site-packages/pip-1.1-py2.7.egg
Processing dependencies for pip
Finished processing dependencies for pip
iMac:pcfb cliburn$ pip install -U xlrd
Downloading/unpacking xlrd from
<SNIP>
Successfully installed xlrd
Cleaning up...
iMac:pcfb cliburn$ pip uninstall xlrd

56

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

Uninstalling xlrd:
/Users/cliburn/Library/Python/2.7/site-packages/xlrd
/Users/cliburn/Library/Python/2.7/site-packages/xlrd-0.7.7-py2.7.egg-info
/Users/cliburn/bin/runxlrd.py
Proceed (y/n)? y
Successfully uninstalled xlrd

4.11.10 Exercise
Write a program that takes a number on the command line and calculates the log, square, sin and cosine, and writes
them out in a csv file.

4.12 NumPy and Matplotlib

4.12.1 NumPy - numeric computing
NumPy is the de facto standard for numerical computing in Python. It is highly optimized and extremely useful
for working with matrices. The standard matrix class in NumPy is called an array. We will first get comfortable
with working with arrays the we will cover a number of useful functions. Then we will touch on the linear algebra
capabilities of NumPy and finally we will use a few examples to tie together key concepts.
NumPy - arrays
Contents

The main object in NumPy is the homogeneous, multidimensional array. An array is a table of elements. An example
is a matrix x

1 2 3
x = 4 5 6
7 8 9
can be represented as
>>> import numpy as np
>>> x = np.array([[1,2,3],[4,5,6],[7,8,9]])
>>> x
array([[1, 2, 3],
[4, 5, 6],
[7, 8, 9]])
>>> x.shape
(3, 3)

The array x has 2 dimensions. In NumPy the number of dimensions is referred to as rank. The ndim is the same as
the number of axes or the length of the output of x.shape
>>> x.ndim
2
>>> x.size
9

Arrays are convenient because of built in methods.

4.12. NumPy and Matplotlib

57

Practical Computing for Biologists, Release 1.0

>>> x.sum(axis=0)
array([12, 15, 18])
>>> x.sum(axis=1)
array([ 6, 15, 24])
>>> x.mean(axis=0)
array([ 4., 5., 6.])
>>> x.mean(axis=1)
array([ 2., 5., 8.])

But arrays are also useful because they interact with other NumPy functions as well as being central to other package
functionality. To make a sequence of numbers, similar to range in the Python standard library, we use arange.
>>> np.arange(10)
array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9])
>>> np.arange(5,10)
array([5, 6, 7, 8, 9])
>>> np.arange(5,10,0.5)
array([ 5. , 5.5, 6. , 6.5, 7. , 7.5,

8. ,

8.5,

9. ,

9.5])

Also we can recreate the first matrix by reshaping the output of arange.
>>> x = np.arange(1,10).reshape(3,3)
>>> x
array([[1, 2, 3],
[4, 5, 6],
[7, 8, 9]])

Another similar function to arange is linspace which fills a vector with evenly spaced variables for a specified interval.
>>> x = np.linspace(0,5,5)
>>> x
array([ 0. , 1.25, 2.5 ,

3.75,

5.

])

As a reminder you may access the documentation at anytime using

~$ pydoc numpy.linspace

Visualizing linspace...
import numpy as np
import matplotlib.pyplot as plt
fig = plt.figure()
ax = fig.add_subplot(111)
N = 8
y = np.zeros(N)
x1 = np.linspace(0, 10, N, endpoint=True)
p1 = plt.plot(x1, y, o)
ax.set_xlim([-0.5,10.5])
plt.show()

58

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

0.06
0.04
0.02
0.00
0.02
0.04
0.06 0

10

Arrays may be made of different types of data.

>>> x = np.array([1,2,3])
>>> x.dtype
dtype(int64)
>>> x = np.array([0.1,0.2,0.3])
>>> x
array([ 0.1, 0.2, 0.3])
>>> x.dtype
dtype(float64)
>>> x = np.array([1,2,3],dtype=float64)
>>> x.dtype
dtype(float64)

There are several convenience functions for making arrays that are worth mentioning:
zeros
ones
>>> x = np.zeros([3,4])
>>> x
array([[ 0., 0., 0., 0.],
[ 0., 0., 0., 0.],
[ 0., 0., 0., 0.]])
>>> x = np.ones([3,4])
>>> x

4.12. NumPy and Matplotlib

59

Practical Computing for Biologists, Release 1.0

array([[ 1.,
[ 1.,
[ 1.,

1.,
1.,
1.,

1.,
1.,
1.,

1.],
1.],
1.]])

Exercise
1. Create the following array (1 line)

1
11

a= .
..

2
12
..
.

..
.

10
20

..
.

91

92

100

2. Use the array object to get the number of elements, rows and columns
3. Get the mean of the rows and columns
4. What do you get when you do this?
>>> a[4,:]

5. [extra] If you have time you can get familiar try

np.log(a)
np.cumsum(a)
np.rank(a)
np.power(a,2)
6. [extra] How do you create a vector that has exactly 50 points and spans the range 11 to 23

NumPy - basics
Quick reference

Here we provide a quick reference guide to the commonly used functions from the NumPy package along with several
frequently encountered examples.
NumPy command
a.ndim
a.shape
arange(start,stop,step)
linspace(start,stop,steps)
dot(a,b)
vstack([a,b])
hstack([a,b])
where(a>x)
argsort(a)

Note
returns the num. of dimensions
returns the num. of rows and colums
returns a sequence vector
returns a evenly spaced sequence in the specificed interval
matrix multiplication
stack arrays a and b vertically
stack arrays a and b horizontally
returns elements from an array depending on condition
returns the sorted indices of an input array

Basic operations

Arithmetic operators in NumPy work elementwise.

60

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

>>> a =
>>> b =
>>> a array([

np.array([3,4,5])
np.ones(3)
b
2., 3., 4.])

Something that can be tricky for people familar with other programming languages is that the * operator does not
carry out a matrix product. This is done with the dot function.
>>> a = np.array([[1,2],[3,4]])
>>> b = np.array([[1,2],[3,4]])
>>> a
array([[1, 2],
[3, 4]])
>>> b
array([[1, 2],
[3, 4]])
>>> a * b
array([[ 1, 4],
[ 9, 16]])
>>> np.dot(a,b)
array([[ 7, 10],
[15, 22]])

Special addition and multiplication operators

>>> a = np.zeros((2,2),dtype=float)
>>> a += 5
>>> a
array([[ 5., 5.],
[ 5., 5.]])
>>> a *= 5
>>> a
array([[ 25., 25.],
[ 25., 25.]])
>>> a + a
array([[ 50., 50.],
[ 50., 50.]])

Concatenation
>>> a = np.array([1,2,3])
>>> b = np.array([4,5,6])
>>> c = np.array([7,8,9])
>>> np.hstack([a,b,c])
array([1, 2, 3, 4, 5, 6, 7, 8, 9])
>>> np.vstack([a,b,c])
array([[1, 2, 3],
[4, 5, 6],
[7, 8, 9]])

Sorting arrays
>>> x.sort()
>>> x
array([0, 0, 1, 3, 4, 5])
>>> x = np.array(([1,3,4,0,0,5]))
array([3, 4, 0, 1, 2, 5])

4.12. NumPy and Matplotlib

61

Practical Computing for Biologists, Release 1.0

>>> np.argsort(x)
array([3, 4, 0, 1, 2, 5])

Common math functions

>>> x = np.arange(1,5)
>>> np.sqrt(x) * np.pi
array([ 3.14159265, 4.44288294,
>>> 2**4
16
>>> np.power(2,4)
16
>>> np.log(np.e)
1.0
>>> x = np.arange(5)
>>> x.max() - x.min()
4

5.44139809,

6.28318531])

Basic operations excercise

Exercise
In the following table we have expression values for 5 genes at 4 time points. These are completely made up data.
Although, some of the questions can be easily answered by looking at the data, microarray data generally come in
much larger tables and if you can figure it out here the same code will work for an entire gene chip.
Gene name
A2M
FOS
BRCA2
CPOX

4h
0.12
0.01
0.03
0.05

12h
0.08
0.07
0.04
0.09

24h
0.06
0.11
0.04
0.11

48h
0.02
0.09
0.02
0.14

1. create a single array for the data (4x4)

2. find the mean expression value per gene
3. find the mean expression value per time point
4. which gene has the maximum mean expression value?
5. sort the gene names by the max expression value

Tip:
>>>
>>>
>>>
>>>
>>>

geneList
values0
values1
values2
values3

=
=
=
=
=

np.array(["A2M", "FOS", "BRCA2","CPOX"])

np.array([0.12,0.08,0.06,0.02])
np.array([0.01,0.07,0.11,0.09])
np.array([0.03,0.04,0.04,0.02])
np.array([0.05,0.09,0.11,0.14])

Additional NumPy

Indexing and Slicing 1D arrays can be indexed in the same way a Python list can.

62

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

>>> a = np.arange(10)
>>> a[2:4]
array([2, 3])
>>> a[:10:2]
array([0, 2, 4, 6, 8])
>>> a[::-1]
array([9, 8, 7, 6, 5, 4, 3, 2, 1, 0])

Multidimensional arrays can have one index per axis

>>> x = np.arange(12).reshape(3,4)
>>> x
array([[ 0, 1, 2, 3],
[ 4, 5, 6, 7],
[ 8, 9, 10, 11]])
>>> x[2,3]
11
>>> x[:,1]
# everything in the second row
array([1, 5, 9])
>>> x[1,:]
# everything in the second column
array([4, 5, 6, 7])
>>> x[1:3,:]
# second and third rows
array([[ 4, 5, 6, 7],
[ 8, 9, 10, 11]])

Where
>>> a = np.array([1,1,1,2,2,2,3,3,3])
>>> a[a>1]
array([2, 2, 2, 3, 3, 3])
>>> a[a==3]
array([3, 3, 3])
>>> np.where(a<3)
(array([0, 1, 2, 3, 4, 5]),)
>>> np.where(a<3)[0]
array([0, 1, 2, 3, 4, 5])
>>> np.where(a>9)
(array([], dtype=int64),)

Printing
>>> for row in x:
...
print row
...
[0 1 2 3]
[4 5 6 7]
[ 8 9 10 11]
>>> for element in x.flat:
...
print element
...
0
1
2
3
4
5
6

4.12. NumPy and Matplotlib

63

Practical Computing for Biologists, Release 1.0

7
8
9
10
11

Copying
>>> a = np.array([a,b,c])
>>> b = a
>>> b[1] = z
>>> a
array([a, z, c],
dtype=|S1)
>>> a = np.array([a,b,c])
>>> b = a.copy()
>>> b[1] = z
>>> a
array([a, b, c],
dtype=|S1)

Missing data
>>> import numpy as np
>>> from scipy.stats import nanmean
>>> a = np.array([[1,2,3],[4,5,np.nan],[7,8,9]])
>>> a
array([[ 1.,
2.,
3.],
[ 4.,
5., nan],
[ 7.,
8.,
9.]])
>>> columnMean = nanmean(a,axis=0)
>>> columnMean
array([ 4., 5., 6.])
>>> rowMean = nanmean(a,axis=1)
>>> rowMean
array([ 2. , 4.5, 8. ])

Generating random numbers

>>> np.random.randint(0,10,5)
# random integers from a closed interval
array([2, 8, 3, 7, 8])
>>> np.random.normal(0,1,5)
# random numbers from a Gaussian
array([ 1.44660159, -0.35625249, -2.09994545, 0.7626487 , 0.36353648])
>>> np.random.uniform(0,2,5)
# random numbers from a uniform distribution
array([ 0.07477679, 0.36409135, 1.42847035, 1.61242304, 0.54228665])

There are many useful functions in random however we are showing only a few so that they will be familar when we
get to plotting.
NumPy - linear algebra
Linear algebra is a branch of mathematics concerned with vector spaces and the mappings between those spaces.
NumPy has a package called linalg. This page is meant only to familiarize you with the NumPys linear algebra
functions for those who are interested.

64

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

A 1 N dimensional vector x

x1
x2

x= .
..
xN
and its transpose xT = (x1 , x2 , . . . , xN ) can be expressed in python as
>>>
>>>
>>>
>>>
(3,
>>>
(1,

import numpy as np
x = np.array([[1,2,3]]).T
xt = x.T
x.shape
1)
xt.shape
3)

A column matrix in NumPy.

3
4

x=
5
6
>>> x = np.array([[3,4,5,6]]).T

A row matrix in NumPy.

5

x= 3

a1,1
a2,1

= .
..

a1,2
a2,2
..
.

..
.

a1,n
a2,n

..
.

am,1

am,2

am,n

>>> x = np.array([[3,4,5,6]])

General matrices you are already familiar with.

Am,n

Common tasks

Matrix determinant
>>> a = np.array([[3,-9],[2,5]])
>>> np.linalg.det(a)
33.000000000000014

Matrix inverse
>>> A = np.array([[-4,-2],[5,5]])
>>> A
array([[-4, -2],
[ 5, 5]])
>>> invA = np.linalg.inv(A)

4.12. NumPy and Matplotlib

65

Practical Computing for Biologists, Release 1.0

>>> invA
array([[-0.5, -0.2],
[ 0.5, 0.4]])
>>> np.round(np.dot(A,invA))
array([[ 1., 0.],
[ 0., 1.]])

Because AA1 = A1 A = I.
Eigenvalues and Eigenvectors
>>> a = np.diag((1, 2, 3))
>>> a
array([[1, 0, 0],
[0, 2, 0],
[0, 0, 3]])
>>> w,v = np.linalg.eig(a)
>>> w;v
array([ 1., 2., 3.])
array([[ 1., 0., 0.],
[ 0., 1., 0.],
[ 0., 0., 1.]])

This is by no means a complete listalso the SciPy package has additional functions if this is an area of interest.
Bibliographic notes

1. Duda, R. O., Hart, P. E. & Stork, D. G. Pattern Classification, John Wiley & Sons, Inc., 2001.
Useful links
NumPy homepage
Official NumPy tutorial
NumPy for MATLAB users

4.12.2 Matplotlib - plotting in Python

Matplotlib - basics
Introduction

The most frequently used plotting package in Python, matplotlib, is written in pure Python and is heavily dependent
on NumPy. The main webpage introduction itemizes what John Hunter (mpl creator) was looking for in a plotting
toolkit.
Plots should look great - publication quality. One important requirement for me is that the text looks good
(antialiased, etc.)
Postscript output for inclusion with TeX documents
Embeddable in a graphical user interface for application development
Code should be easy enough that I can understand it and extend it
66

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

Making plots should be easy

Matplotlib is conceptually divided into three parts:
1. pylab interface (similar to MATLAB) pylab tutorial
2. Matplotlib frontend or API artist tutorial
3. backends drawing devices or renderers
Essentials

The Axes class is the most important class in mpl. The following three lines are used to get an axes class ready for
use.
>>> import matplotlib.pyplot as plt
>>> fig = plt.figure()
>>> ax = fig.add_subplot(2,1,1)

If you need a freehand axis then

>>> fig2 = plt.figure()
>>> ax2 = fig2.add_axes([0.15, 0.1, 0.7, 0.3])

# [left, bottom, width, height]

After a figure is drawn you may save it and or plot it with the following.
>>> fig.saveas(foo.png,dpi=200)
>>> plt.show()

The DPI argument is optional and we can save to a bunch of formats like: JPEG, PNG, TIFF, PDF and EPS.
Here is the example from the artist tutorial.
An example
import numpy as np
import matplotlib.pyplot as plt
fig = plt.figure()
fig.subplots_adjust(top=0.8)
ax1 = fig.add_subplot(211)
ax1.set_ylabel(volts)
ax1.set_title(a sine wave)
t = np.arange(0.0, 1.0, 0.01)
s = np.sin(2*np.pi*t)
line, = ax1.plot(t, s, color=blue, lw=2)
ax2 = fig.add_axes([0.15, 0.1, 0.7, 0.3])
n, bins, patches = ax2.hist(np.random.randn(1000), 50,
facecolor=yellow, edgecolor=yellow)
ax2.set_xlabel(time (s))

4.12. NumPy and Matplotlib

67

Practical Computing for Biologists, Release 1.0

a sine wave

1.0
volts

0.5
0.0
0.5
1.00.0
60
50
40
30
20
10
04

0.2

0.4

0.6

0
time (s)

0.8

1.0

Matplotlib - line and box plots

Simple line plot
Lines with markers
Box plots
Simple line plot
import numpy as np
import matplotlib.pyplot as plt
## initialize the axes
fig = plt.figure()
ax = fig.add_subplot(111)
## format axes
ax.set_ylabel(volts)
ax.set_title(a sine wave)
t = np.arange(0.0, 1.0, 0.01)
s = np.sin(2*np.pi*t)
line, = ax.plot(t, s, color=blue, lw=2)

68

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

a sine wave

1.0

volts

0.5

0.0

0.5

1.00.0

0.2

0.4

0.6

0.8

1.0

Lines with markers

import numpy as np
import matplotlib.pyplot as plt
## initialize the figure
fig = plt.figure()

## the data
t = np.arange(0.0, 1.0, 0.01)
s = np.sin(2*np.pi*t)
## the top axes
ax1 = fig.add_subplot(3,1,1)
ax1.set_ylabel(volts)
ax1.set_title(a sine wave)
line1 = ax1.plot(t, s+5.0, color=blue, lw=2)
line2 = ax1.plot(t, s+2.5, color=red, lw=2)
line3 = ax1.plot(t, s, color=orange, lw=2)
## the middle axes
ax2 = fig.add_subplot(3,1,2)

4.12. NumPy and Matplotlib

69

Practical Computing for Biologists, Release 1.0

ax2.set_ylabel(volts)
ax2.set_title(a sine wave)
line1 = ax2.plot(t, s+5.0, color=black, lw=2,linestyle="--")
line2 = ax2.plot(t, s+2.5, color=black, lw=2,linestyle="-.")
line3 = ax2.plot(t, s, color=#000000, lw=2,linestyle=":")
## the thrid axes
ax3 = fig.add_subplot(3,1,3)
ax3.set_ylabel(volts)
ax3.set_title(a sine wave)
line1 = ax3.plot(t,s+5.0, color=blue, marker="+")
line2 = ax3.plot(t,s+2.5, color=red, marker="o")
line3 = ax3.plot(t,s, color=orange, marker="^")

volts

volts

volts

## adjust the space between plots

plt.subplots_adjust(wspace=0.2,hspace=.4)

6
5
4
3
2
1
0
10.0
6
5
4
3
2
1
0
10.0
6
5
4
3
2
1
0
10.0

a sine wave

0.2

0.4

0.6

0.8

1.0

0.2

0.4

0.6

0.8

1.0

0.2

0.4

0.6

0.8

1.0

a sine wave

a sine wave

Box plots

For more information on box plots try the demo

70

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

import numpy as np
import matplotlib.pyplot as plt

fig = plt.figure()
ax = fig.add_subplot(111)
x1 = np.random.normal(0,1,50)
x2 = np.random.normal(1,1,50)
x3 = np.random.normal(2,1,50)
ax.boxplot([x1,x2,x3])
plt.show()

5
4
3
2
1
0
1
2
3

Matplotlib - bar,scatter and histogram plots

Simple bar plot
Another bar plot
Scatter plot

4.12. NumPy and Matplotlib

71

Practical Computing for Biologists, Release 1.0

Simple bar plot

import numpy as np
import matplotlib.pyplot as plt
fig = plt.figure()
ax = fig.add_subplot(111)
## the data
N = 5
menMeans = [18, 35, 30, 35, 27]
menStd =
[2, 3, 4, 1, 2]
womenMeans = [25, 32, 34, 20, 25]
womenStd =
[3, 5, 2, 3, 3]
## necessary variables
ind = np.arange(N)
width = 0.35

# the x locations for the groups

# the width of the bars

## the bars
rects1 = ax.bar(ind, menMeans, width,
color=black,
yerr=menStd,
error_kw=dict(elinewidth=2,ecolor=red))
rects2 = ax.bar(ind+width, womenMeans, width,
color=red,
yerr=womenStd,
error_kw=dict(elinewidth=2,ecolor=black))
# axes and labels
ax.set_xlim(-width,len(ind)+width)
ax.set_ylim(0,45)
ax.set_ylabel(Scores)
ax.set_title(Scores by group and gender)
xTickMarks = [Group+str(i) for i in range(1,6)]
ax.set_xticks(ind+width)
xtickNames = ax.set_xticklabels(xTickMarks)
plt.setp(xtickNames, rotation=45, fontsize=10)
## add a legend
ax.legend( (rects1[0], rects2[0]), (Men, Women) )
plt.show()

72

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

Scores by group and gender

45

Men
Women

40
35

Scores

30
25
20
15
10

Gr
ou
p5

p4
Gr
ou

Gr

ou
p3

p2
Gr
ou

Gr

ou

p1

Another bar plot

from mpl_toolkits.mplot3d import Axes3D
import matplotlib.pyplot as plt
import numpy as np
fig = plt.figure()
ax = fig.add_subplot(111,projection=3d)
for c, z in zip([r, g, b, y], [30, 20, 10, 0]):
xs = np.arange(20)
ys = np.random.rand(20)
# You can provide either a single color or an array. To demonstrate this,
# the first bar of each set will be colored cyan.
cs = [c] * len(xs)
cs[0] = c
ax.bar(xs, ys, zs=z, zdir=y, color=cs, alpha=0.8)
ax.set_xlabel(X)
ax.set_ylabel(Y)
ax.set_zlabel(Z)
plt.show()

4.12. NumPy and Matplotlib

73

Practical Computing for Biologists, Release 1.0

0.8
0.6 Z
0.4
0.2

10
X

15

20
15
10 Y

25

0.0
30

Scatter plot
#!/usr/bin/env python
import matplotlib.pyplot as plt
import numpy as np
fig = plt.figure()
ax1 = fig.add_subplot(121)
## the data
N=1000
x = np.random.randn(N)
y = np.random.randn(N)
## left panel
ax1.scatter(x,y,color=blue,s=5,edgecolor=none)
ax1.set_aspect(1./ax1.get_data_ratio()) # make axes square

## right panel
ax2 = fig.add_subplot(122)
props = dict(alpha=0.5, edgecolors=none )
handles = []

74

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

colors = [blue, green, magenta, cyan]

for color in colors:
x = np.random.randn(N)
y = np.random.randn(N)
s = np.random.randint(50,200)
handles.append(ax2.scatter(x, y, c=color, s=s, **props))
ax2.set_ylim([-5,11])
ax2.set_xlim([-5,11])
ax2.legend(handles, colors)
ax2.grid(True)
ax2.set_aspect(1./ax2.get_data_ratio())
plt.show()

4
10
blue
3
8
green
2
magenta
6
1
cyan
4
0
2
1
0
2
2
3
4
44 3 2 1 0 1 2 3 4
4 2 0 2 4 6 8 10

Histogram plot

Here is the matplotlib histogram demo

import numpy as np
import matplotlib.pyplot as plt

fig = plt.figure()
ax = fig.add_subplot(111)

4.12. NumPy and Matplotlib

75

Practical Computing for Biologists, Release 1.0

x = np.random.normal(0,1,1000)
numBins = 50
ax.hist(x,numBins,color=green,alpha=0.8)
plt.show()

60
50
40
30
20
10
04

Matplotlib - exercises
Yeast cell cycle data

File download
The data were originally downloaded from the Yeast Cell Cycle Analysis Project Page. These data [1] by Spellman
and Sherlock have likely been used in over a hundred papers.
1. look at the file so you know what you are getting into (less,wc)
2. copy this script into an editor and lets go over it
#!/usr/bin/env python
import csv,os,sys,pickle
import numpy as np
## read the file once to get numRows and numCols
txtFilePath = os.path.join("..","data","cellcycle.txt")

76

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

numRows = 0
reader = csv.reader(open(txtFilePath, r),delimiter=\t)
expListIDs = reader.next()
expListIDs = np.array(expListIDs[1:])
for linja in reader:
numRows+=1
## populate a matrix and name vectors with file info
numColumns = len(expListIDs)
exprMat = np.zeros([numRows,numColumns])
reader = csv.reader(open(txtFilePath, r),delimiter=\t)
header = reader.next()
rowInd = 0
geneList = []
for linja in reader:
row = np.array(linja[1:])
newRow = np.zeros(len(row),)
nanInds = np.where(row == )
goodInds = np.where(row != )
newRow[nanInds] = np.nan
newRow[goodInds] = [float(element) for element in row[goodInds]]
exprMat[rowInd,:] = newRow
rowInd +=1
geneList.append(linja[0])
geneList = np.array(geneList)
## print out info
print ".............."
print "matrix of size (%s,%s) created..."%(exprMat.shape)
print "gene list size - %s"%geneList.size
print "exp list size - %s"%expListIDs.size
print ".............."
## write the data to a file
outFilePath = os.path.join(".","excercise-np.pickle")
tmp = open(outFilePath,w)
pickle.dump([geneList,expListIDs,exprMat],tmp)
tmp.close()

3. save the file using your editor to a directory and use it to read cellcycle.txt. To do this you will need to change
at least one line?
4. create your own script(s) that does the following
opens the pickle file
calculates mean expression value for a gene
plot expression values for a given gene in a histogram
[extra] for a given gene create a lineplot that shows expression values for all the conditions
[extra] make plot that has boxplots for the 5 genes with the greatest expression mean
[extra] create a scatter plot (1 condition) where the negative expression values are green and positive ones
are red
Note that:

4.12. NumPy and Matplotlib

77

Practical Computing for Biologists, Release 1.0

>>> a = np.array([1,2,3,np.nan])
>>> a.max()
nan
>>> a.min()
nan
>>> a.mean()
nan
>>> np.where(np.isnan(a)==False)[0]
array([0, 1, 2])

Also, note that we do not transform, normalize or otherwise process the data in this example. We are using this data
set as a learning tool. The missing data difficulties that arise are common in the biological sciences.
Bibliographic notes

1. Spellman P T, Sherlock, G, Zhang, M Q, Iyer, V R, Anders, K, Eisen, M B, Brown, P O, Botstein, D, Futcher,

B. Comprehensive identification of cell cycle-regulated genes of the yeast Saccharomyces cerevisiae by microarray
hybridization. Molecular biology of the cell, Vol. 9 (12): 3273-97, 1998. PubMed.
Useful links
Matplotlib homepage
Matplotlib gallery

4.13 Biopython I
From the Biopython:
Biopython is a set of freely available tools for biological computation written in Python by an international team of
developers. It is a distributed collaborative effort to develop Python libraries and applications which address the needs
of current and future work in bioinformatics. The source code is made available under the Biopython License, which
is extremely liberal and compatible with almost every license in the world.

4.13.1 Biopython - basics

Introduction
From the biopython website their goal is to make it as easy as possible to use Python for bioinformatics by creating
high-quality, reusable modules and scripts. These modules use the biopython tutorial as a template for what you will
learn here. Here is a list of some of the most common data formats in computational biology that are supported by
biopython.
Uses
Blast
ClustalW
GenBank
PubMed and Medline
ExPASy
SCOP
UniGene
SwissProt
78

Note
finds regions of local similarity between sequences
multiple sequence alignment program
NCBI sequence database
Document database
SIB resource portal (Enzyme and Prosite)
Structural Classification of Proteins (e.g. dom,lin)
computationally identifies transcripts from the same locus
annotated and non-redundant protein sequence database
Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

Some of the other principal functions of biopython.

A standard sequence class that deals with sequences, ids on sequences, and sequence features.
Tools for performing common operations on sequences, such as translation, transcription and weight calculations.
Code to perform classification of data using k Nearest Neighbors, Naive Bayes or Support Vector Machines.
Code for dealing with alignments, including a standard way to create and deal with substitution matrices.
Code making it easy to split up parallelizable tasks into separate processes.
GUI-based programs to do basic sequence manipulations, translations, BLASTing, etc.
Getting started
>>> import Bio
>>> Bio.__version__
1.58

Some examples will also require a working internet connection in order to run.
>>> from Bio.Seq import Seq
>>> my_seq = Seq("AGTACACTGGT")
>>> my_seq
Seq(AGTACACTGGT, Alphabet())
>>> aStringSeq = str(my_seq)
>>> aStringSeq
AGTACACTGGT
>>> my_seq_complement = my_seq.complement()
>>> my_seq_complement
Seq(TCATGTGACCA, Alphabet())
>>> my_seq_reverse = my_seq.reverse()
>>> my_seq_rc = my_seq.reverse_complement()
>>> my_seq_rc
Seq(ACCAGTGTACT, Alphabet())

There is so much more, but first before we get into it we should figure out how to get sequences in and out of python.
File download
FASTA formats are the standard format for storing sequence data. Here is a little reminder about sequences.
Nucleic acid code
A
T
C
G
N
U
D
H

Note
adenosine
thymidine
cytidine
guanine
A/G/C/T (any)
uridine
G/A/T
A/C/T

Nucleic acid code

K
M
R
S
W
B
Y
V

Note
G/T (keto)
A/C (amino)
G/A (purine)
G/C (strong)
A/T (weak)
G/T/C
T/C (pyrimidine)
G/C/A

Here is quickly a bit about how biopython works with sequences

>>> for seq_record in SeqIO.parse(os.path.join("data","ls_orchid.fasta"), "fasta"):
...
print seq_record.id
...
print repr(seq_record.seq)
...
print len(seq_record)
...

4.13. Biopython I

79

Practical Computing for Biologists, Release 1.0

gi|2765658|emb|Z78533.1|CIZ78533
Seq(CGTAACAAGGTTTCCGTAGGTGAACCTGCGGAAGGATCATTGATGAGACCGTGG...CGC, SingleLetterAlphabet())
740

4.13.2 Biopython - sequences and alphabets

The Sequence object
Some examples will also require a working internet connection in order to run.
>>> from Bio.Seq import Seq
>>> from Bio.Alphabet import IUPAC
>>> my_seq = Seq("AGTACACTGGT", IUPAC.unambiguous_dna)
>>> my_seq
Seq(AGTACACTGGT, IUPACUnambiguousDNA())
>>> my_seq.alphabet
IUPACUnambiguousDNA()

A Seq object in python acts like a normal python string.

>>>
...
>>>
>>>
>>>
>>>

for letter in my_seq:

print letter
len(my_seq)
my_seq[4:12]
my_seq[::-1]
str(my_seq)

Nucleotide counts, transcription, translation

>>> my_seq.count("A")
3

to get the GC nucleotide content.

>>> from Bio.SeqUtils import GC
>>> GC(my_seq)
45.45454545454545

transcription and translation

>>> my_mRNA = my_seq.transcribe()
Seq(AGUACACUGGU, IUPACUnambiguousRNA())
>>> my_seq.translate()
Seq(STL, IUPACProtein())

complement and reverse complement

>>> str(my_seq)
AGTACACTGGT
>>> my_seq.complement()
Seq(TCATGTGACCA, IUPACUnambiguousDNA())
>>> my_seq.reverse_complement()
Seq(ACCAGTGTACT, IUPACUnambiguousDNA())

You can translate directly from the DNA coding sequence or you can use the mRNA directly.

80

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

>>> from Bio.Seq import Seq

>>> from Bio.Alphabet import IUPAC
>>> messenger_rna = Seq("AUGGCCAUUGUAAUGGGCCGCUGAAAGGGUGCCCGAUAG", IUPAC.unambiguous_rna)
>>> messenger_rna
Seq(AUGGCCAUUGUAAUGGGCCGCUGAAAGGGUGCCCGAUAG, IUPACUnambiguousRNA())
>>> messenger_rna.translate()
Seq(MAIVMGR*KGAR*, HasStopCodon(IUPACProtein(), *))

Now, you may want to translate the nucleotides up to the first in frame stop codon, and then stop (as happens in nature):
>>> coding_dna.translate()
Seq(MAIVMGR*KGAR*, HasStopCodon(IUPACProtein(), *))
>>> coding_dna.translate(to_stop=True)
Seq(MAIVMGR, IUPACProtein())

Exercise
1. There is so much stuff available in biopython. What happens if you do this?
>>>
>>>
>>>
>>>
>>>

from Bio.Data import CodonTable

standard_table = CodonTable.unambiguous_dna_by_id[1]
mito_table = CodonTable.unambiguous_dna_by_id[2]
print standard_table
print mito_table

The Sequence record object

The SeqRecord objects are the basic data type for the SeqIO objects and they are very similar to Seq objects,however,
there are a few additional attributes.
seq - The sequence itself, typically a Seq object.
id - The primary ID used to identify the sequence a string. In most cases this is something like an accession
number.
name - A common name/id for the sequence a string. In some cases this will be the same as the accession
number, but it could also be a clone name. Analagous to the LOCUS id in a GenBank record.
description - A human readable description or expressive name for the sequence a string.
We can think of the SeqRecord as a container that has the above attributes including the Seq.
Exercise
1. Copy the following script into an editor and save as BioSequences.py
2. Open a terminal window and cd into the appropriate directory.
3. Fill in the missing lines with code
from Bio.Seq import Seq
from Bio.SeqRecord import SeqRecord
## create a simple SeqRecord object
simple_seq = Seq("GATCAGGATTAGGCC")
simple_seq_r = SeqRecord(simple_seq)
simple_seq_r.id = "AC12345"

4.13. Biopython I

81

Practical Computing for Biologists, Release 1.0

simple_seq_r.description = "I am not a real sequence"

## print summary
print simple_seq_r.id
print simple_seq_r.description
print str(simple_seq_r.seq)
print simple_seq_r.seq.alphabet
## translate the sequence
translated_seq = simple_seq_r.seq.translate()
print translated_seq
# exercise 1 -- translate the sequence only until the stop codon

# exercise 2 -- get the reverse complement of the sequence

# exercise 3 -- get the reverse of the sequence (just like for lists)

# exercise 4 -- get the GC nucleotide content

The Sequence IO object

There is one more object that we have to discuss and this the SeqIO object is like a container for multiple SeqRecord
objects.
import os
from Bio import SeqIO

We use a list here to save the gene records from a FASTA file
If you have many records a dictionary will make the program faster.

## save the sequence records to a list

allSeqRecords = []
allSeqIDs
= []
pathToFile = os.path.join("..","data","ls_orchid.fasta")
for seq_record in SeqIO.parse(pathToFile, "fasta"):
allSeqRecords.append(seq_record)
allSeqIDs.append(seq_record.id.split("|")[1])
print seq_record.id
print str(seq_record.seq)
print len(seq_record)
## print out fun stuff about the sequences
print "We found ", len(allSeqIDs), "sequences"
print "information on the third sequence:"
ind = 2
seqRec = allSeqRecords[ind]
print "\t", "GI number
", allSeqIDs[ind]
print "\t", "full id
", seqRec.id
print "\t", "num nucleo.
", len(seqRec.seq)
print "\t", "1st 10 nucleo.", seqRec.seq[:10]

82

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

Just as easy as it is to read a set of files we can save modified versions (i.e. QA). Also, it is almost the exact same code
as above to parse sequences from a GenBank (.gb) file.
There is really way to much to cover in the time we have, but if you have Next Generation Sequencing data then refer
to sections 4.8, 16.1.7 and 16.1.8 of the biopython tutorial. There is even support for binary formats (i.e. SFF).

4.13.3 Bibliographic notes

3. Cock, P J A and Antao, Tiago and Chang, J T and Chapman, B A and Cox, C J and Dalke, A and Friedberg, I
and Hamelryck, T and Kauff, F and Wilczynski B and de Hoon, M J L, Biopython: freely available Python tools
for computational molecular biology and bioinformatics, Bioinformatics, Jun, 2009, 25, 11, 1422-3. PubMed.

4.14 Biopython II
4.14.1 Biopython - Entrez databases
NCBIs Guidelines
Taken from the tutorial.
Before using Biopython to access the NCBIs online resources (via Bio.Entrez or some of the other modules), please
read the NCBIs Entrez User Requirements. If the NCBI finds you are abusing their systems, they can and will ban
your access!
To paraphrase: For any series of more than 100 requests, do this at weekends or outside USA peak times. This is
up to you to obey. Use the http://eutils.ncbi.nlm.nih.gov address, not the standard NCBI Web address. Biopython
uses this web address. Make no more than three requests every seconds (relaxed from at most one request every three
seconds in early 2009). This is automatically enforced by Biopython. Use the optional email parameter so the NCBI
can contact you if there is a problem. You can either explicitly set this as a parameter with each call to Entrez (e.g.
include [email protected] in the argument list), or as of Biopython 1.48, you can set a global email
address:
>>> from Bio import Entrez
>>> Entrez.email = "[email protected]"

Bio.Entrez will then use this email address with each call to Entrez. The example.com address is a reserved domain
name specifically for documentation (RFC 2606). Please DO NOT use a random email its better not to give an
email at all. The email parameter will be mandatory from June 1, 2010. In case of excessive usage, NCBI will attempt
to contact a user at the e-mail address provided prior to blocking access to the E-utilities.
If you are using Biopython within some larger software suite, use the tool parameter to specify this. You can either
explicitly set the tool name as a parameter with each call to Entrez (e.g. include tool=MyLocalScript in the argument
list), or as of Biopython 1.54, you can set a global tool name:
>>> from Bio import Entrez
>>> Entrez.tool = "MyLocalScript"

The tool parameter will default to Biopython. For large queries, the NCBI also recommend using their session history
feature (the WebEnv session cookie string, see Section 8.15). This is only slightly more complicated.

4.14. Biopython II

83

Practical Computing for Biologists, Release 1.0

What databases do I have access to?

>>> from Bio import Entrez
>>> Entrez.email = "[email protected]"
>>> handle = Entrez.einfo()
>>> record = Entrez.read(handle)
>>> record["DbList"]
[pubmed, protein, nuccore, nucleotide, nucgss, nucest, structure, genome,
genomeprj, bioproject, biosample, biosystems, blastdbinfo, books, cdd, clone,
gap, gapplus, dbvar, epigenomics, gene, gds, geo, geoprofiles, homologene,
journals, mesh, ncbisearch, nlmcatalog, omia, omim, pmc, popset, probe,
proteinclusters, pcassay, pccompound, pcsubstance, pubmedhealth, seqannot, snp,
sra, taxonomy, toolkit, toolkitall, unigene, unists, gencoll, gcassembly,
assembly]

What if I want info about a database?

>>> handle = Entrez.einfo(db="pubmed")
>>> record = Entrez.read(handle)
>>> record["DbInfo"]["Description"]
PubMed bibliographic record
>>> record["DbInfo"]["Count"]
21827335

How do I search a db for a given term?

>>> handle = Entrez.esearch(db="pubmed", term="biopython")
>>> record = Entrez.read(handle)
>>> record["IdList"]
[22399473, 21666252, 21210977, 20015970, 19811691, 19773334, 19304878,
18606172, 21585724, 16403221, 16377612, 14871861, 14630660, 12230038]

Other databases?
>>> handle = Entrez.esearch(db="nucleotide",term="Cypripedioideae[Orgn] AND matK[Gene]")
>>> record = Entrez.read(handle)
>>> record["Count"]
75

Get all journals that have computational as a term

>>> handle = Entrez.esearch(db="journals", term="computational")

>>> record = Entrez.read(handle)
>>> record["Count"]
54
>>> record["IdList"]
[39206, 37505, 37511, 37435, 37518, 36366, 39786, 34878, 30367, 38517, 33843,

Ok I have a term now I want the item itself

>>> from Bio import Entrez, SeqIO
>>> handle = Entrez.efetch(db="nucleotide", id="186972394",rettype="gb", retmode="text")
>>> record = SeqIO.read(handle, "genbank")

84

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

>>> handle.close()
>>> print record
ID: EU490707.1
Name: EU490707
Description: Selenipedium aequinoctiale maturase K (matK) gene, partial cds; chloroplast.
Number of features: 3
/sequence_version=1
/source=chloroplast Selenipedium aequinoctiale
/taxonomy=[Eukaryota, Viridiplantae, Streptophyta, Embryophyta, Tracheophyta, Spermatophyt
/keywords=[]
/references=[Reference(title=Phylogenetic utility of ycf1 in orchids: a plastid gene more variable t
/accessions=[EU490707]
/data_file_division=PLN
/date=15-JAN-2009
/organism=Selenipedium aequinoctiale
/gi=186972394
Seq(ATTTTTTACGAACCTGTGGAAATTTTTGGTTATGACAATAAATCTAGTTTAGTA...GAA, IUPACAmbiguousDNA())
>>> handle = Entrez.efetch(db="pubmed", id="21210977")
>>> print handle.read()

4.15 Data management and relational databases

4.15.1 The (very) Basics of SQL
Databases are a very large and complex topic. Classes typically cover weeks, so in our short time we will only scratch
the surface of the basics of selecting, inserting, and joining selects.
Data in relational databases are organized into tables, containing fixed columns of specified data types. For our
example we will use two tables, people and experiment, described below.
create table people (
id integer primary key,
name varchar,
position varchar,
phone varchar,
office varchar
);
create table experiment (
id integer primary key,
name varchar,
researcher integer,
description text,
foreign key(researcher) references people(id)
);

Select
the select command retrieves data from the database.
sqlite> select * from people;
0|Alice|Research Director|555-123-0001|4b
1|Bob|Research assistant|555-123-0002|17

4.15. Data management and relational databases

85

Practical Computing for Biologists, Release 1.0

2|Charles|Research assistant|555-123-0001|24
3|David|Research assistant|555-123-0001|8
sqlite> select * from experiment;
0|EBV Vaccine trial|0|A vaccine trial
1|Flu antibody study|2|Study of the morphology of flu antibodies

The * in the select statement says to select all columns. If you only need a few of the columns you can select them by
name.
sqlite> select name, phone from people;
Alice|555-123-0001
Bob|555-123-0002
Charles|555-123-0001
David|555-123-0001
sqlite> select name, description from experiment;
EBV Vaccine trial|A vaccine trial
Flu antibody study|Study of the morphology of flu antibodies

You can also limit the returned results to rows that match specified information using the where directive.
sqlite> select * from people where name == Alice;
0|Alice|Research Director|555-123-0001|4b
sqlite> select position from people where name == David;
Research assistant

Insert
Adding values to the database is done by using the insert statement.
sqlite> insert into people values ( Null, Edward, Toadie, None, Basement);
sqlite> select * from people where name == Edward;
4|Edward|Toadie|None|Basement

Update
You can also change existing rows once theyve been inserted. update takes a table name as its first argument followed
by set column = value. With out a where clause this will set all rows values. You there for will almost always use the
where clause so that you get specific row/rows values updated.
sqlite> select * from people;
0|Alice|Research Director|555-123-0001|4b
1|Bob|Research assistant|555-123-0002|17
2|Charles|Research assistant|555-123-0001|24
3|David|Research assistant|555-123-0001|8
4|Edward|Toadie|None|Basement
sqlite> update people set name=Eddie where id=4;
sqlite> select * from people;
0|Alice|Research Director|555-123-0001|4b
1|Bob|Research assistant|555-123-0002|17
2|Charles|Research assistant|555-123-0001|24
3|David|Research assistant|555-123-0001|8
4|Eddie|Toadie|None|Basement

86

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

Delete
Similar to updating you can delete rows from the database. The argument again will most likely want a where clause
to prevent deleting all rows in a table.
sqlite> select * from people;
0|Alice|Research Director|555-123-0001|4b
1|Bob|Research assistant|555-123-0002|17
2|Charles|Research assistant|555-123-0001|24
3|David|Research assistant|555-123-0001|8
4|Eddie|Toadie|None|Basement
sqlite> delete from people where name=Eddie;
sqlite> select * from people;
0|Alice|Research Director|555-123-0001|4b
1|Bob|Research assistant|555-123-0002|17
2|Charles|Research assistant|555-123-0001|24
3|David|Research assistant|555-123-0001|8

Joins
The power of relational databases lies in recording relations (the foreign key in the table declaration). To join two
tables you use the join keyword in the select statement and provide a relation to join the two tables. Note, that since
both the people and experiment tables have a column called name we must cast the tables using the as statement.
sqlite> select p.name, e.name from people as p join experiment as e where e.researcher == p.id;
Alice|EPV Vaccine trial
Charles|Flu antibody study

4.15.2 Python and DBI

Working with relational databases is fairly simple with python,
1. Create a connection object
2. Execute a SQL statement
3. Iterate over results
In [1]: import sqlite3
In [2]: con = sqlite3.connect(pcfb.sqlite)
In [3]: r = con.execute(select * from people)
In [4]: for i in r:
...:
print i
(0,
(1,
(2,
(3,
(4,

uAlice, uResearch Director, u555-123-0001, u4b)

uBob, uResearch assistant, u555-123-0002, u17)
uCharles, uResearch assistant, u555-123-0001, u24)
uDavid, uResearch assistant, u555-123-0001, u8)
uEdward, uToadie, uNone, uBasement)

In [5]: r = con.execute(select p.name, e.name from people as p join experiment as e where e.research
In [6]: for i in r:
...:
print Name: %s\n\tExperiment: %s % (i[0],i[1])

4.15. Data management and relational databases

87

Practical Computing for Biologists, Release 1.0

...:
Name: Alice
Experiment: EPV Vaccine trial
Name: Charles
Experiment: Flu antibody study

4.15.3 Exercise:
Write a script to a add a new user and experiment to the database, remove Alice, and reassign her experiments to the
new user. Then have it print out all the experiment names with who owns each experiment.

4.16 Data analysis with Python

We have seen how to perform data munging with regular expressions and Python. For a refresher, here is a Python
program using regular expressions to munge the Ch3observations.txt file that we did on day 1 using TextWrangler.
import re
find = r(\d+)\s+(\w{3})[\w\,\.]*\s+(\d+)\sat\s(\d+):(\d+)\s+([-\d\.]+)\s+([-\d\.]+).*
replace = r\3\t\2.\t\1\t\4\t\5\t\6\t\7
for line in open(examples/Ch3observations.txt):
newline = re.sub(find, replace, line)
print newline,

which when run produces this output

eris:pcfb cliburn$ python examples/regex.py
1752
Jan.
13
13
53
-1.414
1961
Mar.
17
03
46
14
2002
Oct.
1
18
22
36.51
1863
Jul.
20
12
02
1.74

5.781
3.6
-3.4221
133

This session will introduce you to the next stage of data analysis and data visualization. Because it is impossible to do
any justice to these areas in a few hours, the aim of this session is to provide a taste of what analysis and visualization
in Python look like, and a tour of some of the many modules available for scientific computation in Python.

4.16.1 Curve Fitting

One common analysis task performed by biologists is curve fitting. For example, we may want to fit a 4 parameter
logistic (4PL) equation to ELISA data. The usual formula for the 4PL model is
f (x) =

AD
+D
1 + (x/C)B

where x is the concentration, A is the minimum asymptote, B is the steepness, C is the inflection point and D is the
maximum asymptote.
import numpy as np
import numpy.random as npr
import matplotlib.pyplot as plt
from scipy.optimize import leastsq

88

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

def logistic4(x, A, B, C, D):

"""4PL lgoistic equation."""
return ((A-D)/(1.0+((x/C)**B))) + D
def residuals(p, y, x):
"""Deviations of data from fitted 4PL curve"""
A,B,C,D = p
err = y-logistic4(x, A, B, C, D)
return err
def peval(x, p):
"""Evaluated value at x with current parameters."""
A,B,C,D = p
return logistic4(x, A, B, C, D)
# Make up some data for fitting and add noise
# In practice, y_meas would be read in from a file
x = np.linspace(0,20,20)
A,B,C,D = 0.5,2.5,8,7.3
y_true = logistic4(x, A, B, C, D)
y_meas = y_true + 0.2*npr.randn(len(x))
# Initial guess for parameters
p0 = [0, 1, 1, 1]
# Fit equation using least squares optimization
plsq = leastsq(residuals, p0, args=(y_meas, x))
# Plot results
plt.plot(x,peval(x,plsq[0]),x,y_meas,o,x,y_true)
plt.title(Least-squares 4PL fit to noisy data)
plt.legend([Fit, Noisy, True], loc=upper left)
for i, (param, actual, est) in enumerate(zip(ABCD, [A,B,C,D], plsq[0])):
plt.text(10, 3-i*0.5, %s = %.2f, est(%s) = %.2f % (param, actual, param, est))
plt.savefig(logistic.png)

4.16. Data analysis with Python

89

Practical Computing for Biologists, Release 1.0

It will be straightforward to modify this code to use, for example, a five parameter logistic or other equation, offering
a flexibility rarely available with standard analysis software.

4.16.2 Simulation-based statistics

With increasing computational power, it is now feasible to run many, many simulations to estimate parameters instead
of, or in addition to, the traditiional parameteric statistial methods. Most of these methods are based on some form of
resampling of the data available to estimate the null distribution, with well known examples being the bootstrap and
permuation resampling.
Before we do this, we need to understand a little about how to get random numbers. The numpy.random module has
random number generators for a variety of common probabiltiy distributions. These numbers are then used to simulate
the generation of new random samples. If the samples are chosen in a certain way, the statistics of the randomly drawn
samples can provide useful information about the properties of our original data sample. Here are some examples of
random number generation in iptyhon.
In [1]: import numpy.random as npr
In [2]: npr.random(5)
Out[2]: array([ 0.93946009,

0.47060219,

In [3]: npr.random((3,4))
Out[3]:
array([[ 0.29302404, 0.9372624 ,
[ 0.42932372, 0.20717542,
[ 0.83125219, 0.85109042,
In [4]: npr.normal(5, 1, 4)
Out[4]: array([ 3.45966714,

0.10922504,

0.36149538,
0.18447385,
0.32720074,

3.99815107,

0.70776782,

0.41784061])

0.59367473],
0.91159639],
0.33453366]])

5.0191997 ,

5.93739408])

In [5]: npr.randint(1, 7, 10)

Out[5]: array([6, 4, 1, 5, 6, 6, 1, 3, 4, 2])

90

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

In [6]: npr.uniform(1, 7, 10)

Out[6]:
array([ 3.55239006, 5.45862348,
6.23527185, 2.51701757,

1.77176392,
3.96170981,

5.91184323,
6.43878462,

2.49197674,
4.57074553])

In [7]: npr.binomial(n=10, p=0.2, size=(4,4))

Out[7]:
array([[5, 0, 4, 2],
[1, 1, 0, 2],
[3, 2, 0, 2],
[0, 1, 2, 4]])
In [8]: x =

[1,2,3,4,5,6]

In [9]: npr.shuffle(x)
In [10]: x
Out[10]: [2, 1, 5, 6, 4, 3]
In [11]: npr.permutation(10)
Out[11]: array([2, 7, 4, 1, 6, 5, 8, 9, 3, 0])

For example, choosing a new sample with replacement from an existing sample (i.e. we draw one item from the data,
record what it is, then replace it in the data and repeat to get a new sample) can be done efficiently in this way:
In [1]: import numpy as np
In [2]: import numpy.random as npr
In [3]: data = np.array([tom, jerry, mickey, minnie, pocahontas])
In [4]: idx = npr.randint(0, len(data), (4,len(data)))
In [5]: idx
Out[5]:
array([[3, 0,
[4, 2,
[4, 0,
[2, 2,

2,
3,
0,
3,

1,
4,
0,
1,

0],
4],
3],
4]])

In [6]: samples_with_replacement = data[idx]

In [7]: samples_with_replacement
Out[7]:
array([[minnie, tom, mickey, jerry, tom],
[pocahontas, mickey, minnie, pocahontas, pocahontas],
[pocahontas, tom, tom, tom, minnie],
[mickey, mickey, minnie, jerry, pocahontas]],
dtype=|S10)

In the next version of numpy (1.7.0), a new function choice is available in numpy.random to do the same thing
with a nicer syntax. Version 1.7.0 is only currently available from the git repository as source code that you must
compile yourself, but should be available for easy_install/pip installation soon.
In [1]: import numpy.random as npr
In [2]: data = [tom, jerry, mickey, minnie, pocahontas]

4.16. Data analysis with Python

91

Practical Computing for Biologists, Release 1.0

# only availlable if you install numpy 1.7.0 from the git repository
In [3]: npr.choice(data, size=(4, len(data)), replace=True)
--------------------------------------------------------------------------AttributeError
Traceback (most recent call last)
/Volumes/HD3/hg/pcfb/<ipython-input-3-84e7d179a607> in <module>()
----> 1 npr.choice(data, size=(4, len(data)), replace=True)
AttributeError: module object has no attribute choice

Moving on our first simulation example - if we want to plot the 95% confidence interval for the mean of our data
samples, we can use the bootstrap to do so. The basic idea is simple - draw many, many samples with replacement
from the data available, estimate the mean from each sample, then rank order the means to estimate the 2.5 and 97.5
percentile values for 95% confidence interval. Unlike using normal assumptions to calculate 95% CI, the results
generated by the bootstrap are robust even if the underlying data are very far from normal.
import numpy as np
import numpy.random as npr
import pylab
def bootstrap(data, num_samples, statistic, alpha):
"""Returns bootstrap estimate of 100.0*(1-alpha) CI for statistic."""
n = len(data)
idx = npr.randint(0, n, (num_samples, n))
samples = x[idx]
stat = np.sort(statistic(samples, 1))
return (stat[int((alpha/2.0)*num_samples)],
stat[int((1-alpha/2.0)*num_samples)])
if __name__ == __main__:
# data of interest is bimodal and obviously not normal
x = np.concatenate([npr.normal(3, 1, 100), npr.normal(6, 2, 200)])
# find mean 95% CI and 100,000 bootstrap samples
low, high = bootstrap(x, 100000, np.mean, 0.05)
# make plots
pylab.figure(figsize=(8,4))
pylab.subplot(121)
pylab.hist(x, 50, histtype=step)
pylab.title(Historgram of data)
pylab.subplot(122)
pylab.plot([-0.03,0.03], [np.mean(x), np.mean(x)], r, linewidth=2)
pylab.scatter(0.1*(npr.random(len(x))-0.5), x)
pylab.plot([0.19,0.21], [low, low], r, linewidth=2)
pylab.plot([0.19,0.21], [high, high], r, linewidth=2)
pylab.plot([0.2,0.2], [low, high], r, linewidth=2)
pylab.xlim([-0.2, 0.3])
pylab.title(Bootstrap 95% CI for mean)
pylab.savefig(examples/boostrap.png)

92

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

Note that the bootstrap function is a higher order function, and will return the boostrap CI for any valid statistical
function, not just the mean. For example, to find the 95% CI for the standard deviation, we only need to change
np.mean to np.std in the arguments:
# find standard deviation 95% CI bootstrap samples
low, high = bootstrap(x, 100000, np.std, 0.05)

The function is also highly optimized, and takes under 2 seconds to calculate the boostrap mean for a data sample of
size 300 using 100,000 bootstrap samples on a 4 year old MacBook Pro with 2.4 GHz Intel Core 2 Duo processor.
Permutation-resampling is another form of simulation-based statistical calculation, and is often used to evaluate the
p-value for the difference between two groups, under the null hypothesis that the groups are invariant under label
permutation. For example, in a case-control study, it can be used to find the p-value that hypothesis that the mean of
the case group is different from that of the control group, and we cannot use the t-test because the distributions are
highly skewed.
import numpy as np
import numpy.random as npr
import pylab
def permutation_resampling(case, control, num_samples, statistic):
"""Returns p-value that statistic for case is different
from statistc for control."""
observed_diff = abs(statistic(case) - statistic(control))
num_case = len(case)
combined = np.concatenate([case, control])
diffs = []
for i in range(num_samples):
xs = npr.permutation(combined)
diff = np.mean(xs[:num_case]) - np.mean(xs[num_case:])
diffs.append(diff)
pval = (np.sum(diffs > observed_diff) +
np.sum(diffs < -observed_diff))/float(num_samples)
return pval, observed_diff, diffs
if __name__ == __main__:
# make up some data
case = [94, 38, 23, 197, 99, 16, 141]
control = [52, 10, 40, 104, 51, 27, 146, 30, 46]

4.16. Data analysis with Python

93

Practical Computing for Biologists, Release 1.0

# find p-value by permutation resampling

pval, observed_diff, diffs = \
permutation_resampling(case, control, 10000, np.mean)
# make plots
pylab.title(Empirical null distribution for differences in mean)
pylab.hist(diffs, bins=100, histtype=step, normed=True)
pylab.axvline(observed_diff, c=red, label=diff)
pylab.axvline(-observed_diff, c=green, label=-diff)
pylab.text(60, 0.01, p = %.3f % pval, fontsize=16)
pylab.legend()
pylab.savefig(examples/permutation.png)

4.16.3 Data visualization

Data visualization is important for exploratory data analysis as well as for communication of scientific results. Using
ipython with the -- pylab option provides an interarctive environment that is ideal for exploratory data analysis.
The classic Anscombe data set illustrates the importance of visualization when analysing data. The Anscombe data
set consists of 4 different sets of (x,y) values, with esentially identical values for the
1. mean of x
2. variance of x
3. mean of y
4. variance of y
5. correlation between x and y
6. linear regression intercept and slope
Plotting the actual data sets quickly shows that the data sets are not as similar as suggested by the summary stattstics!

94

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

import numpy as np
import pylab
xs = np.loadtxt(anscombe.txt)
for i in range(4):
x = xs[:,i*2]
y = xs[:,i*2+1]
A = np.vstack([x, np.ones(len(x))]).T
m, c = np.linalg.lstsq(A, y)[0]
pylab.subplot(2,2,i+1)
pylab.scatter(x, y)
pylab.plot(x, m*x+c, r)
pylab.axis([2,20,0,14])
pylab.savefig(anscombe.png)

For communication of results, matplotlib offers a huge range of graphics. We have only scratched the surface of
what the package has to offer. The fastest way to get a custom graphic to communicate your results is to look at the
thumbnails at http://matplotlib.sourceforge.net/gallery.html. If one of the graphics looks appropirate for your needs,
just click on the thumbnail to get the source code. You should now know enough Python to customize the graphic to
your specific needs.

4.17 Vector graphics with Inkscape

4.17.1 Graphical concepts
A quote from the text (PCFB).
No longer can you submit a photo that looks right; it has to be a CMYK image, with 300 DPI at printed dimension,
saved as a TIFF with LZW compression. The goal here is to introduce a software that can help with some of these

4.17. Vector graphics with Inkscape

95

Practical Computing for Biologists, Release 1.0

requirements. Please refer to Chapter 17 for further depth than we go into here on graphics.
Vector and pixel images are probably concepts most biologists are familiar with. Pixel images (bitmap, raster) are a
uniform grid of colored dots. In vector art a line can be defined by two endpoints. The vector representation has the
advantage of being often easier to store. Also, the amount of information for vector art stays the same no matter how
large the plot and as you zoom in vector art stays the same.
File formats that store vector based images include: PDF, EPS, SVG and AI. File formats that store pixel-based images
are JPEG, PNG, TIFF, BMP and PSD. PDF, EPS and AI can store embedded pixel information (hybrid images).
It is always possible to go from vector to pixel art, but it is not so easy to go the other way. Another advantage of
vector art is that pixel text is not easily retrieved by a machine.
Pixel art is everywhere (photos, machine output), however if we have to create a completely plot then there are a
number of reasons to do so using vector graphics. Vector art is covered in chapter 18 of the book.

4.17.2 Inkscape
Inkscape uses the SVG (Scalable Vector Graphics) format for its files. SVG is an open standard widely supported by
graphic software.
Inkscape basics
QuickRef

NumPy command
Ctrl+arrow
Ctrl+B
- + =
Ctrl+N
Ctrl+O
Ctrl+S
holding Alt
holding Ctrl
space or F1
[ and ]
arrows
Alt+< and Alt+>
tom
hold Shift + click

Note
pan or scroll the canvas
hides scrollbars
zoom zoom zoom
create a new inkscape documents
opens an exisiting svg file
saves the current file
restricts movement in the move mode
preserves the original height/width ratio during resize
activates the selector
rotate an object
move objects
resize and object. < and > work too
bombadil
selects multiple objects

Getting familiar

1. open a new inkscape document

2. select the square shape and create a square or rectangle by pressing the mouse button once on your canvas
dragging the mouse a little then letting go of the mouse. (try the control squares and circle)
3. Hit the spacebar to go into move/resize mode (try holding Ctrl)
4. Click on the rectangle again to change into rotate/skew mode. (try holding Ctrl) (try [,],<,>)

96

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

Inkscape tutorials
Since there are people that make their living by working with inkscape I figured there were better tutorials already
available on the web than what I could come up with.
Contents

One of the most common issues in inkscape is making arrows. Although, it is a little work once you have made one
then you can save it and use it for all future derivative arrows.
Arrow tutorial
One nice aspect of inkscape is how easy it is to trace a bitmap.

Bitmap tracing tutorial

Try having a look at the Donate button tutorial you can download the svg and open it in inkscape.
Additional tutorials

Now that you are a pro

The coffee cup tutorial
The shiny clock face
Uses
Inkscape can import other formats (e.g. PNG or EPS) then convert.
One could create a image in matplotlib (SVG,PDF etc) and import then annotate

4.17. Vector graphics with Inkscape

97

Practical Computing for Biologists, Release 1.0

Resizing of images
Adding text or shapes or colors to a bitmap
Additional Resources
The inkscape documentation
Nice set of tutorials
Very nice gallery of wallpapers
Python effects tutorial

4.18 Capstone Example

4.18.1 Analysing cytokine array data
In this final example, we will write a program to extract potentially useful summaries of data from a cytokine assay
kindly donated by Herman Staats. You can download the file from the Data Samples page. As you can see, real world
data analysis is far more messy than we have let on so far. Suppose we want to do two things - display the distribution
of each cytokine on diffeerent days to see changes over time, and generate some QC summaries on what errors are
associated with what samples. Doing this manually in Excel will certainly be painful - what if Herman had stashed
away hundreds of such files and wanted to perform the analysis on every single one of them? A manual analysis might
take weeks and will likely be ridden with errors as we grow increasingly bored and frustrated with the task.
So we will write a program to do this instead. The program will be quite challenging to write - we have broken the
task down into subtasks, and your job is to write functions to execute each subtask. Please ask for help if you get stuck
at any point - we recognize that this will be an extemely challenging problem for most of you. OK - lets begin ...
First, we need to convert the spreadsheet into a plain text format by exporting it within Excel. The Data Samples page provides a tab delimited version exported from Excel; exporting to csv format is also fine if you prefer. (Note that Python modules to read and write Excel files are available - see http://www.python-excel.org/ and
http://packages.python.org/openpyxl/ but as these were not covered in the workshop, we will continue to work with
exported plain text files).
Notice that the data is arranged as a nice rectangular grid, with row 1 giving the header information, column 1 the
sample names, and the rest the actual data values or various error messages. This suggests that we can use numpy
arrays to manipulate the data, which will lead to a shorter program than the use of the csv module to parse every item.
Notice also that there is a mixture of strings and numbers in the spreadsheet, and numpy only works with arrays of
all the same type. Since strings can represent numbers but not vice versa, we will first load the data as an array of
strings using the numpy loadtxt function. Our strategy is then to manipulate the data so that apart from the header
(row 1) and sample (column 1) information, everything else is converted to an array of floats for statistical analysis
and visualization.
Our first task is to read the file and convert to an array of strings, which can be done with the loadtxt function. We
first see what loadtxt does:
In [4]: import numpy as np
In [5]: help(np.loadtxt);

98

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

4.18.2 Exercise
Write a function called read_table to load a tab delimited file as a numpy array using the numpy loadtxt
function. The function should take a filename and an optional delimiter (e.g. t for TDL and , for CSV) as arguments,
and return a numpy array of strings:
def read_table(filename, delimiter=\t):
# your code goes here and should return a numpy array of strings

Now that we have the table as a numpy array, lets focus on converting the middle data portion into an array of
floats. We need to somehow convert all those error or warning messages into numbers. One simple way to do thiis is
to encode each unique message as a number that cannot be mistaken for a data value. Since concentrations are never
negative, we will use the negative integers to encode the errors. But in order to do the encoding, we first need to find
out what types of messages exist in the data set.

4.18.3 Exercise
Write a function that finds all the warnings in the table returned by the previous read_table function and returns a
dictionary of {warning : code} where warning is a string (e.g. Bead Issues or UN) and code is a negative number.
Each unique warning should be given a different negative number. Basically we want to find all strings in the data
part of the table, i.e. table[1:, 1:], and everytime we find a string, we shove it into the dictionary if it is not already
there with a new negative number code:
def parse_warnings(table):
# your code goes here and should return a dictionary of warnings

Next, lets deal with the sample information in column 1, excluding row 1, column 1. We notice that the same sample
name can occur on multiple rows - lets create a dictionary whose key is the sample name, and whose value is the row
number minus one (minus one because we want row number = 0 to index the first data row, not the header information).

4.18.4 Exercise
Complete the function:
def parse_samples(table):
# your code goes here and should return a dictionary of lists of row numbers for each sample

Parsing the header information is slightly more challenging. We notice that two different pieces of information are
provided by each cell in the header - the cytokine name (e.g. IL-2 beta) and the day the sample was taken (e.g. day 3).
Each cytokine is sampled on multple days (0, 3 and 28). We want to create a dictionary that will tell us what column
we can find the values for a given cytokine, day combination. One way to deal with the fact that the same cytokine is
sampled on multiple days is to use a dictionary of dictionaries. In particular, we will construct a dictionary whose key
is a cytokine name, and whose value is another dictionary. This other dictionary has a key representing the day, and a
value representing the column number minus one for the (cytokine, day) combination.

4.18.5 Exercise
Complete the function:
def parse_headers(table):
# your code goes here and should return a dictionary of dictionaries as described in the text

We finally get to convert the data from an array of strings to an array of floats. Replace all the warning messages with
the code numbers, and return a new array of floats comprising the following subarray - table[1:, 1:].

4.18. Capstone Example

99

Practical Computing for Biologists, Release 1.0

4.18.6 Exercise
Complete the function:

def parse_data(table, warning_dict):

# your code goes here and should return an array of floats containing values for the entries in tab

Lets consolidate all the above functions into a single function that when given a filename, reads the file and breaks it
down into a sample dictionary, a cytokine dicitonary and a data array. My function is shown below - you only have to
change the names to match the structures that you created:
def parse_cytokine_table(filename):
# extract data and sample and cytokine mapping dictinaries from table
table = read_table(filename)
warning_dict = parse_warnings(table)
sample_mapper = parse_samples(table)
cytokine_mapper = parse_headers(table)
data = parse_data(table, warning_dict)
return warning_dict, sample_mapper, cytokine_mapper, data

Well done! The most difficult part of the program is now complete. We next see how we can use the structures we
have created to make summaries of the data. First, lets use a box-and-whiskers plot to show the distribution of each
cytokine over days 0, 3 and 28. Generate one such figure for each such cytokine - it should look something like this.

4.18.7 Exercise
Write the function to generate the box-and -whiskers plots. Remember that the negative numbers are not really
cytokine concentrations and should not be used for plotting:
def plot_cytokine(cytokine, cytokine_mapper, data, save=False, directory=.):
# generate plots like the one shown

100

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

# if save is False, dont save but simply show each figure

# if save if True, save to disk with savefig(directory/filename)

Finally, lets write the QC data to file. There seem to be an enormous number of errors in this file - lets summarize
them so that Herman can figure out what is going on! Generate one text file per unique warning, where each row has
two columns - the sample name, and the number of that type of warning associated with the sample. For instance, the
Bead_Issues.txt file will contain the following values:
0921-X-2-2
1059-X-2-2
0273-X-2-2
0740-X-2-2
0263-X-2-2
0175-X-2-2
0012-X-2-2
0108-X-2-2
0066-X-2-2
0057-X-2-2
0313-X-2-2
0103-X-2-2
0685-X-2-2
0799-X-2-2
0749-X-2-2
0693-X-2-2
1023-X-2-2
0300-X-2-2

2
12
2
4
18
2
4
1
2
12
7
25
2
4
16
1
6
2

Complete the function:

def write_qc(qc, directory=.):
# write QC reports as described
# save to the directory given and use the mesage name + .txt as the filename

That wraps up the coding part of this final capstone example.

4.18.8 Final Exercise

Now that you know how to think like a programmer, dont create such spreadsheets in the future if possible! Design
spreadsheets so that the information is machine-friendly. What are some of the ways you can make the original
spreadsheet more machine-friendly if you could change the design?
The full program that for parsing the cytokne assay file:
1
2
3
4
5

"""Intermediate
Task: Summarize
Input: An Excel
Output: Figures
"""

example of data munging using numpy.

the change in concentration over time for each cytokine.
worksheet exported as a tab delimited file
summarizing concentrations of individual cytokines over time.

6
7
8
9

import numpy
import pylab
import os

10
11
12
13

def parse_cytokine_table(filename):
# extract data and sample and cytokine mapping dictinaries from table
table = read_table(filename)

14
15

warning_dict = parse_warnings(table)

4.18. Capstone Example

101

Practical Computing for Biologists, Release 1.0

sample_mapper = parse_samples(table)
cytokine_mapper = parse_headers(table)
data = parse_data(table, warning_dict)
return warning_dict, sample_mapper, cytokine_mapper, data

16
17
18
19
20
21
22
23
24
25

def read_table(filename, delimiter=\t):

# read data into an array of strings
# this can also be done by standard data munging techniques
# but would be more painful
return numpy.loadtxt(filename, dtype=string, delimiter=delimiter)

26
27
28
29
30
31
32
33
34
35
36
37
38
39

def parse_warnings(table):
# create a dictionary to convert missing data codes into numeric codes
data = table[1:, 1:]
warning_dict = {}
idx = -1
for x in data.ravel():
try:
float(x)
except:
if x not in warning_dict:
warning_dict[x] = idx
idx -= 1
return warning_dict

40
41
42
43
44
45
46
47

def parse_samples(table):
# create a dictionary that returns the rows where each sample has data
samples = table[1:,0]
sample_mapper = {}
for i, sample in enumerate(samples):
sample_mapper.setdefault(sample, []).append(i)
return sample_mapper

48
49
50
51
52
53
54
55
56
57

def parse_headers(table):
# get header information in first row
headers = table[0,:]
# parse header informtion to find columns for each cytokine/day combination
cytokine_mapper = {}
for i, label in enumerate(headers[1:]):
cytokine, day = label.split(day)
cytokine_mapper.setdefault(cytokine.strip(), {})[int(day)] = i
return cytokine_mapper

58
59
60
61
62
63
64
65

def parse_data(table, warning_dict):

# convert missing data from strings to code numbers for analysis
data = table[1:, 1:]
for warning in warning_dict:
data[data==warning] = warning_dict[warning]
data = data.astype(float)
return data

66
67
68
69
70

def plot_cytokine(cytokine, cytokine_mapper, data, save=False, directory=.):

# generate box and whiskers plot for distribution of a single cytokine
idx_dict = cytokine_mapper[cytokine]
xs = data[:, sorted(idx_dict.values())]

71

# collect measured values for each day in a list ignoring missing values
ys = []

72
73

102

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

Practical Computing for Biologists, Release 1.0

74
75
76

for i in range(xs.shape[1]):
col = xs[:,i]
ys.append(col[col >= 0])

77
78
79
80
81
82
83
84

# if there are any measured values, visualize the data disbution

if numpy.sum(len(y) for y in ys) > 0:
pylab.figure()
pylab.boxplot(ys)
pylab.xticks(range(1, len(idx_dict.keys())+1), idx_dict.keys())
pylab.xlabel(Days)
pylab.title(cytokine)

85
86
87
88
89

if save:
if not os.path.exists(directory):
os.makedirs(directory)
pylab.savefig(os.path.join(directory, cytokine + .png))

90
91
92
93
94
95
96
97
98
99
100
101

def sample_qc(warning_dict, sample_mapper, data):

# generate a dictionary whose keys are warning labels
# and whose values are lists of (sample, number of warnings) pairs
qc = {}
for warning in warning_dict:
for sample in sample_mapper:
idx = sample_mapper[sample]
xs = data[idx, :]
num_warnings = numpy.sum(xs==warning_dict[warning])
qc.setdefault(warning, []).append((sample, num_warnings))
return qc

102
103
104
105

def write_qc(qc, directory=.):

if not os.path.exists(directory):
os.makedirs(directory)

106
107
108
109
110
111
112
113
114

for k in sorted(qc):
filename = k.replace( , _).replace(/, -) + .txt
fo = open(os.path.join(directory, filename), w)
sample_list = qc[k]
for sample in sample_list:
if sample[1] != 0:
fo.write(%s\t%d % sample + \n)
fo.close()

115
116
117
118
119

if __name__ == __main__:
# parse the file
warning_dict, sample_mapper, cytokine_mapper, data = parse_cytokine_table(
Cytokine_assay_31Dec08PAD.txt)

120
121
122
123
124

# visualize cytokine distributions

for cytokine in cytokine_mapper:
plot_cytokine(cytokine, cytokine_mapper, data,
save=True, directory=figures)

125
126
127
128

# Save errors by sample for QC to output files

qc = sample_qc(warning_dict, sample_mapper, data)
write_qc(qc, directory=qc_reports)

4.18. Capstone Example

103

Practical Computing for Biologists, Release 1.0

104

Chapter 4. Instructor: Cliburn Chan, Biostatistics and Bioinformatics.

INDEX

A
assignments, 9

P
participants, 11

R
references, 10

105