PepMLM: Target Sequence-Conditioned Generation of Therapeutic Peptide Binders via Span Masked Language Modeling

2023/10/05 by Tianlai Chen, Madeleine Dumas, Chen, Tianlai +38 · 3 voices · 5 citations
Biochemistry, Genetics and Molecular Biology · Medicine · #Peptidase Inhibition and Analysis #Protein Degradation and Inhibitors #q-bio.BM #vaccines and immunoinformatics approaches

paper · pdf · doi:10.48550/arxiv.2310.03842

openalex publication_date 2023/10/05 · openalex created_date 2023/10/25 · openalex updated_date 2026/08/01

Abstract

benchmarking with AlphaFold-Multimer, outperforming RFDiffusion on structured targets, we experimentally verify PepMLM's efficacy via fusion of model-derived peptides to E3 ubiquitin ligase domains, demonstrating endogenous degradation of emergent viral phosphoproteins and Huntington's disease-driving proteins. In total, PepMLM enables the generative design of candidate binders to any target protein, without the requirement of target structure, empowering downstream therapeutic applications.

Citations

• Unconstrained genome targeting with near-PAMless engineered CRISPR-Cas9 variants
• Biological structure and function emerge from scaling unsupervised learning to 250 million protein sequences
• ProtTrans: Toward Understanding the Language of Life Through Self-Supervised Learning
• Highly accurate protein structure prediction with AlphaFold
• PROTAC targeted protein degraders: the past is prologue
• Large language models generate functional protein sequences across diverse families
• Evolutionary-scale prediction of atomic-level protein structure with a language model
• De novo design of protein structure and function with RFdiffusion

Cited by

• Generative design and validation of therapeutic peptides for glioblastoma based on a potential target ATP5A
• Full-Atom Peptide Design via Riemannian-Euclidean Bayesian Flow Networks
• EnzyControl: Adding Functional and Substrate-Specific Control for Enzyme Backbone Generation
• Sequence-based protein-protein interaction prediction and its applications in drug discovery
• NLP4Neuro: Sequence-to-sequence learning for neural population decoding

Discussions

• “PepMLM: Target Sequence-Conditioned Generation of Peptide Binders via Masked Language Modeling” 🧶🧬 Fine-tunes ESM-2 network to achieve “target-conditioned de novo binder design from sequence alone [bsky, 6 points, 0 comments]
• Happy to share our early work on generating binding peptides conditioned ONLY on the target sequence! 🌟 PepMLM masks cognate peptides at the end of target protein sequences, and tasks ESM-2 to fully [bsky, 5 points, 0 comments]
• PepMLM: Target Sequence-Conditioned Generation of Peptide Binders via Masked Language Modeling arxiv.org/abs/2310.0... --- #proteomics #prot-preprint [bsky, 0 points, 0 comments]

Related

• ApexGen: Simultaneous design of peptide binder sequence and structure for target proteins
• Prot42: a Novel Family of Protein Language Models for Target-aware Protein Binder Generation
• Computational design of target-specific linear peptide binders with TransformerBeta
• PepDoRA: A Unified Peptide Language Model via Weight-Decomposed Low-Rank Adaptation
• Hotspot-Driven Peptide Design via Multi-Fragment Autoregressive Extension