Deep Origin

We're nearing the end of pre-registration for the #DeepOriginChallenge with a few spots left! To recap, the Deep Origin challenge: 🌄 Is a day-long AI/ML for drug discovery hackathon (though no life science experience is needed) 🤝 Lets you bring a team of up to 4 (or sign up yourself) 🤖 Let's you meet and collaborate with leading ML and drug discovery experts in Yerevan 💵 Compete for cash prizes Grab one of the few remaining spots below! https://lnkd.in/gew5-er3

We all know that more data is needed to build more predictive models in drug development. But generating data is costly, including from quantum mechanical simulations such as Density Functional Theory (DFT) calculations, which provides valuable information on the shapes and electron distributions of small molecules and informs how they interact with protein targets. Existing DFT datasets also contain many molecules that just don't look like drugs. To help with this, our team at Deep Origin developed the Smart Distributed Data Factory, a volunteer computing platform designed to crowdsource DFT calculations -- think SETI@home but for molecules! The model uses machine learning to identify which molecules in a set will be most informative for training DFT models, then distributes them into volunteers' computers, using about ~10 minutes of compute to create a new data point. Using SDDF, we created a dataset of 2 million DFT calculations for molecules in the Enamine Ltd. REAL DB, freely available to anyone. More broadly, we hope the SDDF framework will be useful to other researchers to crowdsource simulation dataset generation in the drug discovery space. Check out the paper here 👉  https://lnkd.in/gEfih6R5 To download the DFT dataset or volunteer compute to create new DFT data points for the community, head here 👉  https://sddfactory.cloud/

One of our favorite things about #DeepOriginChallenge is seeing the range of talent and expertise in AI, math, and programming in #Yerevan. Know a student, student, or drug hunter who wants to join this year’s challenge? Tag them below or sign up here 👉https://lnkd.in/gew5-er3

We're now accepting applications for this year's #DeepOriginChallenge! You can sign up as teams of 1-4 to spend a day at our #Yerevan office collaborating on interesting machine learning problems for a chance to win $1,000's. No biotech domain expertise needed and we promise it'll be a good time! Sign up here 👉 https://lnkd.in/gJUeWGfx Fun fact: this year's challenge is organized by last year's participants who are already part of our growing team. #Hackathon #AIML #Biotech #DrugDiscovery

New Episode: ‘How we built Oculus VR’ In this latest episode of The Humanware Project, I'm thrilled to share my captivating conversation with Michael Antonov, co-founder of Oculus and now CEO of Deep Origin. Michael reveals how a 19-year-old's DIY virtual reality headset grew into Oculus, a company that Facebook would later acquire for $2 billion. From late-night coding sessions to breakthrough moments in positional tracking, he shares the behind-the-scenes story of VR's revival and what it takes to build something truly revolutionary. In this wide-ranging discussion, Michael shares: ✨ The pivotal moment when Oculus caught John Carmack's attention and set VR's comeback in motion 👨🦳 Why he believes ageing is the biggest technological challenge facing humanity 🧑🔬 How Deep Origin is working to democratise drug discovery through better computational tools 🩺 His practical approach to health optimisation and thoughts on biohacking 🙌 Why success often comes from pursuing what genuinely excites you. This conversation isn't just about technology – it's about the mindset that drives innovation and the courage to tackle seemingly impossible challenges. Whether you're an aspiring entrepreneur, a technology enthusiast, or simply curious about the future of human health and longevity, Michael's insights offer a fascinating glimpse into what's possible when vision meets persistence. 🎧 Watch the full episode now: https://lnkd.in/e5yp4WdA I'd love to hear your thoughts! And who should we interview next? #WellnessTechnology #Oculus #VR #Ageing #Healthcare #Innovation #TheHumanWareProject

It’s been incredible visiting New York City and having the opportunity to host Deep Origin event alongside Nucleate New York and Merck Digital Sciences Studio - MDSS with Ahmet Hamdi C. Beyond this event, Nucleate New York also co-hosted another fantastic gathering with External Affairs, Genspace NYC, Columbia Alumni Association Global Health & Wellness Network, and The Innovation Space - Applied Vol 8 For those who haven’t heard about Deep Origin yet… The team is developing Balto, an AI-driven, physics-based computational platform designed for molecular modeling, cheminformatics, in silico drug design, and more. 🔬 Key capabilities of Balto (https://lnkd.in/g5QxDRw4) : 🧬 Molecular Data Retrieval – Integrated access to ChEMBL, PDB, UniProt, PubChem, and more. ⚖ Molecular Property & ADMET Predictions – LogP, LogS, hERG inhibition, CYP450 interactions, etc. 🛠 Protein Structure & Ligand Prep – Pocket identification, protonation state prediction, molecular similarity search. 🎯 Molecular Docking & Screening – AI-powered docking workflows for hit identification. 📊 Computational Data Processing – Automated dataset handling and workflow optimization. 👀 Molecular Visualization – Interactive structure rendering and binding pocket analysis. Beyond Balto’s core tools, Deep Origin also supports molecular dynamics, free energy perturbation (FEP) calculations, large-scale virtual screening, and more through enterprise collaborations. It was amazing meeting so many biotech professionals from the NYC ecosystem. I’m looking forward to staying in touch and seeing what exciting collaborations come next! Natalie Ma, PhD Steve Cheney Joseph Borrello Jacob Trevino Julia Nanyi Zhao, PhD #DeepOrigin #Biotech #AIforDrugDiscovery #ComputationalBiology #MolecularModeling #NucleateNYC #MerckDigitalSciencesStudio #Networking #DrugDiscovery #Startups #NYC

It was a pleasure to showcase Deep Origin's tools and capabilities last week at the converge meet up hosted by Biocom California and Bits in Bio. Natalie Ma, PhD. gave an excellent demo of Balto, showcasing its capabilities in examining molecular structures, exploring databases, analyzing binding poses, conducting molecular docking simulations, and serving as a highly convenient, ChatGPT-like tool for various protein and small molecule-related tasks. Working on Drug Discovery and want a trial? Just send a DM. Also wonderful listening to some of the brilliant efforts in the field presented by: Nikhil Patel Product Manager at Chan Zuckerberg Initiative - and learn about the new developments towards creation of their whole cell model. Karmen Čondić-Jurkić, CEO at Open Molecular Software Foundation. - Herwork in creating open source data sets for a spectrum of functions is commendable. If you are looking for open source data sets and communities- you should talk to them. Samantha Liang, Director of Alliances and Collaborations at Parker Institute for Cancer Immunotherapy. I cant wait for their data sets BRUCE and RADIOHEAD to be public. Something to keep an eye out for if you are in Oncology. BRUCE: Multi-omic data sets of patients treated with Immunotherapy. RADIOHEAD: Multi-omic datasets generated from patients receiving checkpoint inhibitors. The session was fantastically moderated by Nicholas Larus-Stone. Thanks to Michelle Nemits from Biocom California for organising this wonderful event hosted at PwC. Also loved the after party at Mithrl, thanks for hosting: Vivek Adarsh and Michael Linsner. . . . #biotech #drugdiscovery #oncology #cancer #bitsinbio #drugdevelopment #Sanfrancisco #machinelearning

The #DeepOriginChallenge is back! Join us in our Yerevan office for a day of competition and collaboration building efficient #machineLearning models in small teams. Prizes up to $3k and the promise to meet interesting and innovative leaders in AI and drug discovery. Note: no chemistry domain knowledge is required. Though our challenge will revolve around #molecularModeling with AI. Register below!

Aram Davtyan giving an awesome poster presentation about our work modeling protein-small molecule-protein ternary complexes at Keystone Symposia on Molecular & Cellular Biology Proximity-Based Therapeutics! #biotech #drugDev #degraders #molecularGlues #PROTACs Check out the poster yourself here 👇 https://lnkd.in/diUPp--B

Another incredible Converge Meet-up by Biocom California! It’s always a pleasure working with Biocom California, and I’m grateful Nucleate Bay Area for co-hosting, as well as PwC for their sponsorship and support. The event was fantastic - amazing presentations, great discussions, wonderful people, and a truly beautiful space! A special highlight was Natalie Ma, PhD presentation on Deep Origin tools. She gave an excellent demo of Balto, showing how it can be used daily in the lab, whether for quickly checking molecular structures, searching databases, analyzing binding poses, doing molecular docking simulations, or just having a super-convenient “ChatGPT” like tool for any proteins and small molecules tasks. Make sure to check the 1-month free trial today! Thank you to everyone who made this event a success! Looking forward to the next one. #BiocomCalifornia #NucleateBayArea #ConvergeMeetup #DeepOrigin #Balto #Biotech #DrugDiscovery #AIinBiotech