We scan new podcasts and send you the top 5 insights daily.
Current co-folding and structural models memorize ligand-pocket pairings rather than understanding binding physics. When binding pockets are digitally mutated, filled, or deleted, structural models frequently predict that small molecules still bind in the non-existent pocket because of memorized positions.
To evolve AI from pattern matching to understanding physics for protein engineering, structural data is insufficient. Models need physical parameters like Gibbs free energy (delta-G), obtainable from affinity measurements, to become truly predictive and transformative for therapeutic development.
Attention maps in binding models initially focus on non-functional, variable amino acid regions outside active domains, essentially creating protein nicknames to bypass learning physics. Screening systematically mutated targets with disrupted pockets forces cross-attention models to distribute focus evenly across actual binding residues.
The community standard of two-angstrom accuracy for protein-ligand predictions is insufficient. At that resolution, critical details like an aromatic ring's orientation can be wrong, rendering the model's output misleading for drug design. Genesis argues one-angstrom accuracy is the minimum for practical utility.
Models like AlphaFold don't solve protein folding from physics alone. They heavily rely on co-evolutionary data, where correlated mutations across species provide strong hints about which amino acids are physically close. This dramatically constrains the search space for the final structure.
An anecdote about a "wonky" BindCraft design with disconnected beta sheets, which experts predicted would fail, highlights a key trend. The resulting binder was one of the best ever produced, suggesting AI models are extracting structural principles that go beyond traditional human "protein literacy" and intuition.
Current tools excel at the static binding problem. To advance to creating true therapeutics, models must incorporate the physics of displacing solvent and ions from an interface—currently neglected but one of the "biggest enemies" of strong binding in a physiological context.
Current AI for protein engineering relies on small public datasets like the PDB (~10,000 structures), causing models to "hallucinate" or default to known examples. This data bottleneck, orders of magnitude smaller than data used for LLMs, hinders the development of novel therapeutics.
AlphaFold reproduced static structures deposited in the Protein Data Bank rather than modeling complete protein dynamics or actual conformational ground states. Proteins are complex, disordered, and context-dependent, meaning biological understanding of dynamics remains an open challenge.
Standard co-folding models struggle with induced proximity because public databases like the PDB lack sufficient training data for multi-protein complexes. Because dynamic multi-protein complexes take on endless conformations, standard AI models become overconfident and inaccurate. To solve this, Protai generates proprietary structural proteomics data that maps physical distances and conformations, enabling accurate modeling of complexes that can be targeted by molecular glues and PROTACs.
Generative AI alone designs proteins that look correct on paper but often fail in the lab. DenovAI adds a physics layer to simulate molecular dynamics—the "jiggling and wiggling"—which weeds out false positives by modeling how proteins actually interact in the real world.