Apheris @ PEGS EU

​The evening before PEGS Europe, we're bringing a small group of large-molecule people together at Contra, down at Doca de Santo Amaro, for drinks, food and scientific conversations.
Most of the antibody data that could make models better sits inside individual companies and never gets shared. We run federated data networks, where pharma companies train models together on their proprietary data without it ever leaving their own environment.

​In AISB-1, the first initiative of our AI Structural Biology (AISB) Network, training on more than 20,000 private structures from five pharma companies raised the share of high-accuracy protein-protein interface predictions (PPI-lDDT ≥ 0.8) from 39% to 63%. We're now extending this approach to antibodies. Join to learn more and take an active role in shaping this collaborative journey with us.

​Robin Röhm (chairing Wednesday's PEGS roundtable on federated collaboration) and Leonardo Castorina will host for Apheris.