In this work, we introduce PepMLM, a purely target sequence-conditioned de novo generator of linear peptide binders. By employing a novel masking strategy that uniquely positions cognate peptide sequences at the terminus of target protein sequences, PepMLM tasks the state-of-the-art ESM-2 pLM to fully reconstruct the binder region, achieving low perplexities matching or improving upon previously-validated peptide-protein sequence pairs. After successful in silico benchmarking with AlphaFold-Multimer, we experimentally verify PepMLM's efficacy via fusion of model-derived peptides to E3 ubiquitin ligase domains, demonstrating endogenous degradation of target substrates in cellular models. In total, PepMLM enables the generative design of candidate binders to any target protein, without the requirement of target structure, empowering downstream programmable proteome editing applications.
Check out our manuscript on the arXiv!
# Load model directly
from transformers import AutoTokenizer, AutoModelForMaskedLM
tokenizer = AutoTokenizer.from_pretrained("TianlaiChen/PepMLM-650M")
model = AutoModelForMaskedLM.from_pretrained("TianlaiChen/PepMLM-650M")
For a faster response, please tag either @Leo-T-Zang or @pranamanam. Thank you!
@article{chen2025target,
title={Target sequence-conditioned design of peptide binders using masked language modeling},
author={Chen, Leo Tianlai and Quinn, Zachary and Dumas, Madeleine and Peng, Christina and Hong, Lauren and Lopez-Gonzalez, Moises and Mestre, Alexander and Watson, Rio and Vincoff, Sophia and Zhao, Lin and others},
journal={Nature Biotechnology},
pages={1--9},
year={2025},
publisher={Nature Publishing Group US New York}
}
