Language Modeling Materializes a World Model of Protein Biology
- 1Biohub
- 2work performed at Biohub or EvolutionaryScale
Abstract
Proteins are fundamental to life. The full extent of their biology is beyond our ability to characterize with experimental approaches in the physical laboratory. Accurate digital representations could accelerate the discovery of protein biology through virtual experiments. We propose language modeling to learn unified and general representations that can be scaled to all of protein biology. Building on these representations, we develop a structure prediction model that exceeds the performance of established methods for biomolecular complex prediction across benchmarks, including for the interactions of antibodies with their targets. A simple search procedure yields high experimental success rates for the discovery of proteins with nanomolar binding affinities for both miniproteins and single-chain antibodies, a modality critical for therapeutic design. Study of the concepts in the language model’s representation space reveals a systematic organization aligned with the reductionist understanding of proteins developed through empirical science. Leveraging this organization, we generate a comprehensive map of protein biology encompassing over 6.8 billion sequences and 1.1 billion predicted structures, identifying connections across known and unknown biology. As a whole, this shows language modeling as a powerful substrate for representing the biology of proteins, operating across scales from the prediction and design of protein interactions at the atomic level, to identifying properties of proteins at different levels of granularity and abstraction, to the scale of mapping connections between proteins across billions of years of evolution.
Related
- LEMON-ZEST: Evolution-Informed Tokenization for Efficient Protein Language Modeling
- Targeted protein degradation: bridging chemical biology and clinical translation
- Advances in computational biology approaches for digital modeling of tumor microenvironment.
- Peptidomics and Computational Docking Prioritize Candidate Peptide–Protein Associations Relevant to Upper-Gastrointestinal Biology in a Fermented Soy-Derived Postbiotic
- Understanding language model scaling for protein fitness prediction
BibTeX
@misc{candido2026language,
title = {{Language Modeling Materializes a World Model of Protein Biology}},
author = {Salvatore Candido and Thomas Hayes and Alexander Derry and Roshan Rao and Zeming Lin and Robert Verkuil and Bryan Z. Wu and Jin Sub Lee and Elise S. Bruguera and Jehan A. Keval and Mykhailo Kopylov and John E. Pak and Wesley C. H. Wu and Neil Thomas and Samson J. Mataraso and Alvin Hsu and Ashton Trotman‐Grant and Kilian Fatras and Allan dos Santos Costa and Rohil Badkundri and Halil Akin and Deniz Oktay and Jonathan Deaton and Elizabeth A. Montabana and Hrishita Sitwala and Yue Yu and Marius Wiggert and Dylan Alexander Carlin and Anthony W. Goering and Tomasz Blazejewski and McCullen Sandora and Michael Hla and Tina Z. Jia and Leon H. Kloker and Nicholas Sofroniew and Masatoshi Uehara and Jaspreet Pannu and Sharrol T. Bachas and Daniel S. Liu and Tom Sercu and Alexander W. Rives},
howpublished = {bioRxiv (Cold Spring Harbor Laboratory)},
year = {2026},
month = jun,
doi = {10.64898/2026.06.03.729735},
url = {https://doi.org/10.64898/2026.06.03.729735},
note = {CC BY},
}
RIS
TY - UNPB
TI - Language Modeling Materializes a World Model of Protein Biology
AU - Salvatore Candido
AU - Thomas Hayes
AU - Alexander Derry
AU - Roshan Rao
AU - Zeming Lin
AU - Robert Verkuil
AU - Bryan Z. Wu
AU - Jin Sub Lee
AU - Elise S. Bruguera
AU - Jehan A. Keval
AU - Mykhailo Kopylov
AU - John E. Pak
AU - Wesley C. H. Wu
AU - Neil Thomas
AU - Samson J. Mataraso
AU - Alvin Hsu
AU - Ashton Trotman‐Grant
AU - Kilian Fatras
AU - Allan dos Santos Costa
AU - Rohil Badkundri
AU - Halil Akin
AU - Deniz Oktay
AU - Jonathan Deaton
AU - Elizabeth A. Montabana
AU - Hrishita Sitwala
AU - Yue Yu
AU - Marius Wiggert
AU - Dylan Alexander Carlin
AU - Anthony W. Goering
AU - Tomasz Blazejewski
AU - McCullen Sandora
AU - Michael Hla
AU - Tina Z. Jia
AU - Leon H. Kloker
AU - Nicholas Sofroniew
AU - Masatoshi Uehara
AU - Jaspreet Pannu
AU - Sharrol T. Bachas
AU - Daniel S. Liu
AU - Tom Sercu
AU - Alexander W. Rives
PY - 2026
DA - 2026/06/04
PB - bioRxiv (Cold Spring Harbor Laboratory)
DO - 10.64898/2026.06.03.729735
UR - https://doi.org/10.64898/2026.06.03.729735
AB - Proteins are fundamental to life. The full extent of their biology is beyond our ability to characterize with experimental approaches in the physical laboratory. Accurate digital representations could accelerate the discovery of protein biology through virtual experiments. We propose language modeling to learn unified and general representations that can be scaled to all of protein biology. Building on these representations, we develop a structure prediction model that exceeds the performance of established methods for biomolecular complex prediction across benchmarks, including for the interactions of antibodies with their targets. A simple search procedure yields high experimental success rates for the discovery of proteins with nanomolar binding affinities for both miniproteins and single-chain antibodies, a modality critical for therapeutic design. Study of the concepts in the language model’s representation space reveals a systematic organization aligned with the reductionist understanding of proteins developed through empirical science. Leveraging this organization, we generate a comprehensive map of protein biology encompassing over 6.8 billion sequences and 1.1 billion predicted structures, identifying connections across known and unknown biology. As a whole, this shows language modeling as a powerful substrate for representing the biology of proteins, operating across scales from the prediction and design of protein interactions at the atomic level, to identifying properties of proteins at different levels of granularity and abstraction, to the scale of mapping connections between proteins across billions of years of evolution.
ER -