’ Correspondence

Artificial intelligence (AI) is paving the way for

a critical role in drug discovery, drug design,
a+kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC1y0abggQZXdgGj2MwlZLeI= on 11/07/2024

and studying drug–drug interactions –

Downloaded from http://journals.lww.com/international-journal-of-surgery by BhDMf5ePHKav1zEoum1tQfN4

correspondence
Sandip Chakraborty, MVSca, Hitesh Chopra, PhDb, Shopnil Akash, MPharme,*, Chiranjib Chakraborty, PhDd,
Kuldeep Dhama, MVSc, PhDc,*

Dear Editor, marketing allocate resources for maximum share gain in the market
In the pharmaceutical sector, there has been a sudden accelera- and reverse poor sales[2].
tion in the digitalization of data over the past few years. This Various case studies have identified the potential use of AI in
digitalization has come with challenges to acquire, scrutinize, and the discovery of drugs. AI is being successfully used for the
apply such particular knowledge for reaching solutions to complex identification of novel compounds for treating cancer.
clinical problems, which has ultimately motivated applications of Researchers have trained a deep learning (DL) algorithm upon a
artificial intelligence (AI) as it is able to handle huge volumes of data huge dataset of known compounds related to cancer and their
with augmentation of automation. AI is a system based on tech- corresponding biological activity. Novel compounds with greater
nology involving several modern tools and networks that mimic the potential for the treatment of cancer have been obtained as an
intelligence of humans. Side by side, there is no threat of the phy- output. This demonstrates the capability of the method for dis-
sical presence of humans being completely replaced by AI. Potential covering new candidates for therapy. Recently, the utility of ML
applications of AI are being extended continuously in the field of for the identification of small molecule inhibitors of the mitogen-
pharmaceuticals, and developing pharmaceutical products from the activated protein kinase/ERK kinase (MEK) 22 has been descri-
bench to the side of the bed is imaginably provided[1]. AI can help in bed. MEK is also a target for treating cancer, but developing
efficacious inhibitors is a challenge for scientists. In this context,
designing drugs rationally, assist in decision-making, help in the
the identification of novel inhibitors of this protein by the ML
determination of the correct therapeutic management of patients,
algorithm has great clinical significance. Another example is the
and be wisely exploited for developing drugs in the future. Eularis
use of an ML algorithm to identify novel inhibitors of beta-
has developed an AI platform for analysis and making decisions
secretase (BACE1), a protein found in relation to the development
known as E-VAI (Eularis Virtual Analytics Interface) that uses
of Alzheimer’s disease. AI has also been successfully used for the
machine learning (ML) algorithms along with an easily usable user discovery of new antibiotics; a powerful ML-based approach has
interface for creating analytical roadmaps. This helps executives in identified powerful antibiotics from a vast pool of 10 crore
molecules and more, including one that acts against several
a
Department of Veterinary Microbiology, College of Veterinary Sciences and Animal bacteria. For example, the drug halicin can destroy
Husbandry, West Tripura, Tripura, bDepartment of Biosciences, Saveetha School of
Mycobacterium tuberculosis and carbapenem-resistant strains of
engineering, Saveetha Institute of Medical and Technical Sciences, Chennai, Tamil
Nadu, cDivision of Pathology, ICAR-Indian Veterinary Research Institute, Bareilly, the family Enterobacteriaceae[3–8]. The discovery of drug mole-
Izatnagar, Uttar Pradesh, dDepartment of Biotechnology, School of Life Science and cules like DSP-1181 and ISM001-055 have entered phase 1
Biotechnology, Adamas University, Kolkata, West Bengal, India and eDepartment of clinical trials for treating patients with obsessive-compulsive
Pharmacy, Faculty of Allied Health Science, Daffodil International University, Daffodil disorder (OCD) and idiopathic pulmonary fibrosis (IPF),
Smart City, Ashulia, Savar, Dhaka, Bangladesh
respectively, in humans reveals the promise that AI holds for the
Sponsorships or competing interests that may be relevant to content are disclosed at
the end of this article.
future discovery of many more drug molecules[9,10].
Now let us deeply analyze the role of AI in the drug discovery
*Corresponding author. Address: Department of Pharmacy, Faculty of Allied
Health Sciences, Daffodil International University, Dhaka 1207, Bangladesh. process. For the prediction of the desired chemical structure of a
Tel.: + 880 193 556 7417. E-mail: [email protected] (S. Akash); compound, several parameters, viz., predictive models, molecular
Division of Pathology, Indian Veterinary Research Institute, Izatnagar 243122, Bareilly, similarity, the process of generation of molecules, and the utiliza-
Uttar Pradesh, 243122, India. Tel.: +91 581 231 0074/+91 983 765 4996; tion of approaches in silico, can be used. A new system known as
fax: +91 581 230 3284. E-mail: [email protected] (K. Dhama).
DeepVS has been presented for docking several receptors and
Copyright © 2023 The Author(s). Published by Wolters Kluwer Health, Inc. This is an
ligands and has shown excellent performance. Another approach
open access article distributed under the terms of the Creative Commons
Attribution-Non Commercial-No Derivatives License 4.0 (CCBY-NC-ND), where it is has used a multiobjective replacement algorithm (automated) for
permissible to download and share the work provided it is properly cited. The work optimizing the potency profile of an inhibitor of cyclin-dependent
cannot be changed in any way or used commercially without permission from the kinase-2 by assessing its similarity in shape, biochemical action, and
journal.
physiochemical features[11,12]. Quantitative structure–activity rela-
International Journal of Surgery (2023) 109:3242–3244 tionship (QSAR) modeling tools are used to identify potential
Received 3 June 2023; Accepted 10 June 2023 candidate drugs and have gradually developed into QSAR (AI-
Published online 21 June 2023 based) approaches. Such approaches include linear discriminant
http://dx.doi.org/10.1097/JS9.0000000000000564 analysis (LDA), support vector machines (SVMs), random forests

3242
 Correspondence. International Journal of Surgery (2023)

(RF), and decision trees. All these can be used to speed up QSAR integration of experimental methods available traditionally with
analysis[2]. For predicting the three-dimensional (3D) pharmaco- AI hold promise to overcome the limitations as well as challenges
phore features, implementation of ML technology has been done of AI in the process of drug discovery. The ever-growing attention
by HS-Pharm, Pharm-IF, and DeepSite. Two models have been and interest of researchers and scientists, pharmaceutical com-
built by Coveney’s group for binding affinities with accuracy and panies, and regulatory agencies, in combination with the benefits
reproducibility and have been implemented in ML. The models are of AI, potentially make it an attractive research area that holds
thermodynamic integration with enhanced sampling (TIES) and great promise. Ultimately, it can be said that AI has the potential
a+kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC1y0abggQZXdgGj2MwlZLeI= on 11/07/2024

enhanced sampling of molecular dynamics with approximation of to transform the process of drug discovery in modern times.
continuum solvent (ESMACS). For the prediction of binding affi-
Downloaded from http://journals.lww.com/international-journal-of-surgery by BhDMf5ePHKav1zEoum1tQfN4

nities after the step of molecular docking, eSimDock is another tool.

Improvement of the accuracy of ligand ranking is done by inte- Ethical approval
grating scoring functions by the use of ML (nonlinear). The No ethical approval is required for these types of articles.
insensitivity of eSimDock to deformations or distortions in the
target structure makes it a tool with flexibility that can be used
irrespective of the protein structure’s quality[13]. Consent
Proposals have been put forward for several neural network
algorithms for designing drugs. Such algorithms include simple Not required.
ones like multilayer perceptron (MLP) on the basis of
McCulloch–Pitts neurons or more complex ones like logistic, Sources of funding
naïve Bayes, ridge, lasso, shallow neural networks, or SVMs.
Satisfactory results have been shown by tree-based models among No funding was received.
the techniques applied for designing drugs. There are wide uses of
decision trees in problems that concern prediction as well as the
classification of concepts and ideas. There is involvement in the Author contribution
construction of multiple decision trees in a random forest. S.C. and H.C.: conceptualization and writing original draft; S.A.,
Random forests are actually ML algorithms commonly used that C.C., and K.D.: review and editing.
combine the output of more than one decision tree to reach to
only one result. The most complex DL network models include
generative adversarial networks (GANs), convolutional neural Conflicts of interest disclosure
networks (CNNs), and capsule networks (CapsNets)[14]. The
GANs play a crucial role in generating helical protein (novel) The authors declare no conflicts of interest.
backbone topologies, which is important from the perspective of
designing proteins de novo. Moreover, the GAN framework can Research registration unique identifying number
be implemented for producing new peptides for drug targets[15]. (UIN)
Drug–drug interaction (DDI) can be studied by applying learn-
ing-based methods like CNN-DDI[16]. For the development of Not applicable.
classification models of hERG blockers and non-blockers,
CapsNets has been adopted in a study by Wang et al.[17], wherein
the excellent performance of CapsNets has proven their potential Guarantor
applications in studies related to drug discovery. Shopnil Akash.
A hierarchical DL framework is RPITER, which includes CNN
and stacked autoencoders. Prediction of the interactions of the
protein with ribonucleic acid (RNA) (non-coding) can be done Provenance and peer review
with a fair degree of accuracy by using RPITER. Further
employment of probabilistic matrix factorization has been done Not commissioned, externally peer-reviewed.
to predict the interaction between the drug and target. This
method is suitable to be applied to a large volume of data, as for Data availability statement
predictions to be made accurately, the method requires a plethora
of data inputs. The development of pairwise input neural net- No data sets were generated during this study.
works (multichannel) has also been done for predicting
compound–protein interactions, benefiting from the limited
amount of data available in the literature[13,18,19]. References
In conclusion, AI has a great role to play in revolutionizing the [1] Chopra H, Baig AA, Gautam RK, et al. Application of artificial intelli-
gence in drug discovery. Curr Pharm Des 2022;28:2690–703.
process of drug discovery and can offer better accuracy and
[2] Paul D, Sanap G, Shenoy S, et al. Artificial intelligence in drug discovery
efficiency. AI can also accelerate the process of drug development and development. Drug Discov Today 2021;26:80–93.
and has the ability to develop more efficacious and personalized [3] Marchant J. Powerful antibiotics discovered using AI. Nature 2020.
medicines and therapies. But it should not be forgotten that the doi:10.1038/d41586-020-00018-3
success of AI in the discovery of high-quality drugs is dependent [4] Melo MCR, Maasch JRMA, de la Fuente-Nunez C. Accelerating antibiotic
discovery through artificial intelligence. Commun Biol 2021;4:1050.
on the availability of data. Moreover, ethical concerns need to be [5] Gupta R, Srivastava D, Sahu M, et al. Artificial intelligence to deep
addressed appropriately. Recent developments in the field of AI learning: machine intelligence approach for drug discovery. Mol Divers
technology, viz, augmentation of data, explainable AI, and 2021;25:1315–60.

3243
 Correspondence. International Journal of Surgery (2023) International Journal of Surgery

[6] Zhu J, Wang J, Wang X, et al. Prediction of drug efficacy from [13] Arabi AA. Artificial intelligence in drug design: algorithms, applications,
transcriptional profiles with deep learning. Nat Biotechnol 2021;39: challenges and ethics. Future Drug Discov 2021;3:FDD59.
1444–52. [14] Staszak M, Staszak K, Wieszczycka K, et al. Machine learning in drug
[7] Dhamodharan G, Mohan CG. Machine learning models for predicting design: Use of artificial intelligence to explore the chemical structure–
the activity of AChE and BACE1 dual inhibitors for the treatment of biological activity relationship. Wiley Interdiscip Rev Comput Mol Sci
Alzheimer’s disease. Mol Divers 2022;26:1501–17. 2022;12:e1568.
[8] The Guardian. Powerful antibiotic discovered using machine learning for [15] Lin E, Lin CH, Lane HY. Relevant applications of generative adversarial
first time | Antibiotics | The Guardian. [Accessed 3 June 2023]. https:// networks in drug design and discovery: molecular de novo design,
a+kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC1y0abggQZXdgGj2MwlZLeI= on 11/07/2024

www.theguardian.com/society/2020/feb/20/antibiotic-that-kills-drug- dimensionality reduction, and de novo peptide and protein design.

resistant-bacteria-discovered-through-ai Molecules 2020;25:3250.
Downloaded from http://journals.lww.com/international-journal-of-surgery by BhDMf5ePHKav1zEoum1tQfN4

[9] BBC News. Artificial intelligence-created medicine to be used on humans [16] Zhang C, Lu Y, Zang T. CNN-DDI: a learning-based method for pre-
for first time – BBC News. [Accessed 3 June 2023]. https://www.bbc.com/ dicting drug–drug interactions using convolution neural networks. BMC
news/technology-51315462 Bioinform 2022;23:88.
[10] Forbes. First Wholly AI-Developed Drug Enters Phase 1 Trials. [Accessed [17] Wang Y, Huang L, Jiang S, et al. Capsule networks showed excellent
3 June 2023]. https://www.forbes.com/sites/calumchace/2022/02/25/ performance in the classification of hERG blockers/nonblockers. Front
first-wholly-ai-developed-drug-enters-phase-1-trials/?sh=585f647c2680 Pharmacol 2020;10:1631.
[11] Firth NC, Atrash B, Brown N, et al. MOARF, an integrated workflow for [18] Lee M, Kim H, Joe H, et al. Multi-channel PINN: investigating scalable
multiobjective optimization: implementation, synthesis, and biological and transferable neural networks for drug discovery. J Cheminform
evaluation. J Chem Inf Model 2015;55:1169–80. 2019;11:46.
[12] Pereira JC, Caffarena ER, Dos Santos CN. Boosting docking-based virtual [19] Peng C, Han S, Zhang H, et al. Rpiter: a hierarchical deep learning frame-
screening with deep learning. J Chem Inf Model 2016;56:2495–506. work for ncRNA–protein interaction prediction. Int J Mol Sci 2019;20:1070.

3244