BAGEL-CAR: Reflections on the Bits to Binders Competition
The antigen-binding segment of chimeric antigen receptors (CARs) in CAR-T therapy has emerged as a compelling application of de novo AI protein design. In the Bits to Binders competition, our group submitted 414 BAGEL-CAR designs for CD20-directed CAR binding segments, 38.4% of which were statistically enriched in a pooled CAR-T proliferation screen, the highest among all teams and more than…
The Bits to Binders competition showcased a groundbreaking application of de novo AI protein design in the realm of CAR-T therapy. Our team submitted a substantial number of BAGEL-CAR designs, accounting for 38.4% of the total entries. These designs demonstrated remarkable success, with 38.4% of them emerging statistically enriched in a proliferation screen, a feat unmatched by any other team.
Two distinct design strategies were found to be the driving force behind our success, outperforming all competing teams in aggregate results. Notably, we refrained from applying a conventional validation step known as ESMFold refolding, which could have potentially disqualified all our experimentally enriched designs. This finding underscores the importance of stress-testing each filtering step per biomolecular system and assessing intermediate outputs of design pipelines, regardless of prevailing practices.
Our research suggests that designing against ESMFold's predicted CD20 dimer geometry, which significantly differs from the experimentally resolved 6Y97 structure, may have played a pivotal role in our high hit rate. By sharing these insights, we aim to contribute to the ongoing discourse on the evolution of CAR-T and binder design campaigns.
Written by urgent.news from bioRxiv's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
