Validation and analysis of 12,000 AI-driven CAR-T designs in the Bits to Binders competition

Abstract Artificial intelligence (AI) methods for proteins have advanced rapidly, improving structure prediction and design, particularly for de novo binders. However, most evaluations emphasize binding affinity rather than higher-order biological function. We present Bits to Binders , a global competition evaluating de novo binder design in the context of chimeric antigen receptor (CAR) T cells. Teams from 42 countries submitted 12,000 designs of 80-amino acid de novo minibinders targeting human CD20 as CAR binding domains. Designs were screened by pooled CAR-T proliferation, identifying 707 designs exhibiting significant CD20-specific enrichment, with team hit rates from 0.6% to 38.4%. Top-performing candidates were validated as individual constructs, measuring CD20-specific proliferation, expansion, cytokine production, and targeted cell lysis. We identified common design methodologies and factors correlated with DNA synthesis, expression, and target-specific T cell activation which nearly double the success rates when applied as a retrospective filter. We release this dataset as an open resource, with practical recommendations to support more effective AI-driven binder design.

Authors

Publication Details

Journal
Molecular Systems Biology
Published
2026-10-08
DOI
https://doi.org/10.1038/s44320-026-00246-1
Citations
1
Primary Topic
CAR-T cell therapy research
Type
article
Field-Weighted Citation Impact
2.78
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

Validation and analysis of 12,000 AI-driven CAR-T designs in the Bits to Binders competition

Moritz Ertelt, Vivian Haas, Jokent T. Gaza, Fatemeh Nasiri et al.
1 citations
Molecular Systems Biology
CAR-T cell therapy research
2.78
article

Validation and analysis of 12,000 AI-driven CAR-T designs in the Bits to Binders competition

Moritz Ertelt, Vivian Haas, Jokent T. Gaza, Fatemeh Nasiri, Clayton W. Kosonocky, Aaron L. Feller, Alia Clark‐ElSayed, Javier Marchena-Hurtado, Tyler Derr, Jimin Pei, Kaveh Nasrollahi, Qian Cong, Stefano Angioletti‐Uberti, Tjaša Mlakar, Ajasja Ljubetič, Aoi Otani, Tanner Dean, Roman Jerala, Patrick Bryant, Qiuzhen Li, Parth Bibekar, Rocco Moretti, Seyed Alireza Hashemi, Akshay Chenna, Jie Chen, Rohit Satija, Phillip R. Woolley, Jesse Durham, Delbert S. Barth, Rajat Punia, Duško Lainšček, Jakub Lála, Anton Bushuiev, Nicole Chiang, Max Beining, Abel Gurung, Elliot Cole, Daksh Joshi, Alex M. Abel, ASHISH MAKANI, Marko Ludaic, Aswini Javvadi, Amanda E Cifuentes Rieffer, Jens Meiler, Max 1888-1959 Lingner, Sai Jaideep Reddy Mure, Pragati Naikare, Alberto Perez, Bits to Binders Competitors, Calvin XiaoYang Hu, Haelyn Kim, Hakyung Lee, Corey Howe, Dieter Hoffmann, Diego Kleiman, Saeideh Moradvandi, Joseph Clark, David Miller, Nuria Mitjavila, Francisco Requena, Fernando Meireles, Ansar Ahmad Javed, Petr Kouba, Ava Chan, Karl Lundquist, Rana A Barghout, Huan Koh, Vivian Chu, Daniel Acosta, Yong Youn Kwon, Jinling Huang, Sarah Knapp, Alejandro Diaz, Joseph Openy, Dionessa Biton, Dominic Rieger, Ora Furman, Elias Sanchez, Vinayak Annapure, Neil Anthony, Yuxuan Liu, Aparna Sahu, Yisel Martinez Noa, Tynan Gardner, Shreyasi Das, Qianchen Liu, Nader Ibrahim, Rui Guo, Freddie Martin, Vikrant Parmar, Sofiia Hoian, Satoshi Ishida, Nathaniel Greenwood, Cianna Calia, Yunchao Liu, Even Kiely, Antonio Fonseca, Rajarshi Mondal
article en
1 citations

Abstract

Abstract Artificial intelligence (AI) methods for proteins have advanced rapidly, improving structure prediction and design, particularly for de novo binders. However, most evaluations emphasize binding affinity rather than higher-order biological function. We present Bits to Binders , a global competition evaluating de novo binder design in the context of chimeric antigen receptor (CAR) T cells. Teams from 42 countries submitted 12,000 designs of 80-amino acid de novo minibinders targeting human CD20 as CAR binding domains. Designs were screened by pooled CAR-T proliferation, identifying 707 designs exhibiting significant CD20-specific enrichment, with team hit rates from 0.6% to 38.4%. Top-performing candidates were validated as individual constructs, measuring CD20-specific proliferation, expansion, cytokine production, and targeted cell lysis. We identified common design methodologies and factors correlated with DNA synthesis, expression, and target-specific T cell activation which nearly double the success rates when applied as a retrospective filter. We release this dataset as an open resource, with practical recommendations to support more effective AI-driven binder design.

Molecular Systems Biology
Openalex Percentile: Top 6%
CAR-T cell therapy research
2.78
DOIOpen Access PDFOpenAlex
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.