Open App
4 hours
The Hangar
Free Tickets Available
Mon, 19 May, 2025 at 04:00 pm to 08:00 pm (GMT-07:00)
The Hangar
460 Forbes Boulevard, South San Francisco, United States
From Models to Molecules: AI's Expanding Roles in Therapeutics
Join us at The Hangar, right outside Genentech, on May 19 (Monday) afternoon, for an exciting event exploring the intersection of artificial intelligence and machine learning in the field of biotherapeutics. Dive deep into how these technologies are revolutionizing the way we develop and optimize treatments. Connect with experts, network with like-minded individuals, and gain insights into the future of healthcare. Don't miss out on this opportunity to learn and grow in a dynamic environment!
Info: Google DeepMind amazed the world with the release of AlphaFold, promising a revolution in how therapeutics are discovered and developed. However, five years on, the number of AI-designed drugs that have reached the clinic remains surprisingly small. Despite massive investment and ambitious claims, the impact of AI in therapeutic discovery is still unfolding—and proving more complex than many expected. Here we discuss some of the critical challenges that have slowed progress: data quality and scarcity, the unpredictable nature of biological systems, and the gap between in silico predictions and clinical success. While the hype has often outpaced the results, we should remain optimistic about what AI can deliver. To overcome these obstacles, we at 310 AI are building an AI platform purpose-built for drug discovery, designed to generate insights that translate into actionable results. By rethinking how AI is applied in this domain, we aim to help realize its true potential and impact.
Info: The discovery of therapeutic protein biologics has been limited by two main challenges: the difficulty of reasoning about protein–protein interactions and the need to co-optimize multiple drug-like traits—such as solubility, stability, and immunogenicity—that often conflict with binding affinity. Aikium’s Yotta-ML² platform addresses these barriers by uniting massive-scale wet lab screening with deep learning–driven design. Its core, Yotta-Display, screens trillion-member libraries—100–1,000× larger than phage display and ~100,000× that of yeast—unlocking vast sequence and structural diversity. Generating over 10 million labeled data points weekly, Yotta-ML² enables precise modeling of protein interfaces and prediction of key biophysical traits. We use it to design selective binders against some of the toughest drug targets, including intrinsically disordered regions of membrane proteins and transcription factors—advancing a new era of precision biologics for “undruggable” targets.
Info: Antibody-drug conjugates (ADCs) are powerful cancer therapies, but optimizing pharmacokinetics, especially clearance and half-life, remains challenging. We present a framework that integrates AlphaFold Multimer (AF-M) and machine learning (ML) to predict ADC half-life using detailed molecular descriptors. AF-M enables accurate antibody structure prediction, allowing extraction of structural features like hydrophobicity and charge distributions. These, combined with ADC-specific attributes—drug-to-antibody ratio, conjugation sites, linker type, payload properties—train robust ML models. Using 118 antibodies and ADCs in mice, our cross-validated XGBoost models achieved a Pearson correlation of 0.76 between predicted and experimental half-life, outperforming sequence-based models. Time-split tests confirmed strong predictive performance within the model’s domain. This work highlights the value of combining AlphaFold-derived features with ML for improved ADC design and PK optimization.
Info: Deep Origin is building a virtual cell, a predictive in silico system of biological models to predict drug safety and efficacy. Using a mix of AI and mechanistic methods their models have outperformed Schrodinger, AlphaFold, and other state-of-the-art models. They have developed DO-AWSEM, a protein dynamics model that combines physics and neural networks to enable efficient simulation of protein folding, protein-protein interaction, and protein macromolecule interaction to enable better understanding of antibody-antigen interactions and prioritize candidates.
Also check out other Business events in South San Francisco, Arts events in South San Francisco.
Tickets for From Models to Molecules: AI's Expanding Roles in Therapeutics can be booked here.
| Ticket type | Ticket price |
|---|---|
| Biotech/Pharma professionals & researchers | Free |
| Investors | Free |
| Academic researchers | Free |
We have gathered all the information for you in one convenient spot, but please keep in mind that these are subject to change.We do our best to keep everything updated, but something might be out of sync. For the latest updates, always check the official event details by clicking the "Find Tickets" button.
