paper-with-me

홈 › Papers

Improving Inverse Folding for Peptide Design with Diversity-regularized Direct Preference Optimization

2024-10-25 · Ryan Park, Darren J. Hsu, C. Brian Roland, Maria Korshunova, Chen Tessler, Shie Mannor, Olivia Viessmann, Bruno Trentini

Inverse folding models play an important role in structure-based design by predicting amino acid sequences that fold into desired reference structures. Models like ProteinMPNN, a message-passing encoder-decoder model, are trained to reliably produce new sequences from a reference structure. However, when applied to peptides, these models are prone to generating repetitive sequences that do not fold into the reference structure. To address this, we fine-tune ProteinMPNN to produce diverse and structurally consistent peptide sequences via Direct Preference Optimization (DPO). We derive two enhancements to DPO: online diversity regularization and domain-specific priors. Additionally, we develop a new understanding on improving diversity in decoder models. When conditioned on OpenFold generated structures, our fine-tuned models achieve state-of-the-art structural similarity scores, improving base ProteinMPNN by at least 8%. Compared to standard DPO, our regularized method achieves up to 20% higher sequence diversity with no loss in structural similarity score.

📄 PDF Abstract BibTeX arXiv:2410.19471

Code (0)

등록된 구현이 없습니다.

Tasks

DecoderDiversity

Methods 이 논문이 사용한 방법론

BASE 설명 없음
DPO 설명 없음

Similar Papers 제목 키워드 기반

AlphaFold Distillation for Protein Design

2022-10-05 · Igor Melnyk, Aurelie Lozano, Payel Das, Vijil Chenthamarakshan

Inverse protein folding, the process of designing sequences that fold into a specific 3D structure, is crucial in bio-engineering and drug discovery. Traditional methods rely on experimentally resolved structures, but th…

DiversityDrug DiscoveryKnowledge DistillationModel Optimization+2

Reinforcement learning on structure-conditioned categorical diffusion for protein inverse folding

2024-10-22 · Yasha Ektefaie, Olivia Viessmann, Siddharth Narayanan, Drew Dresser 외

Protein inverse folding-that is, predicting an amino acid sequence that will fold into the desired 3D structure-is an important problem for structure-based protein design. Machine learning based methods for inverse foldi…

DiversityProtein Design

MuCO: Generative Peptide Cyclization Empowered by Multi-stage Conformation Optimization

2026-01-30 · Yitian Wang, Fanmeng Wang, Angxiao Yue, Wentao Guo 외 arxiv

Modeling peptide cyclization is critical for the virtual screening of candidate peptides with desirable physical and pharmaceutical properties. This task is challenging because a cyclic peptide often exhibits diverse, ri…

Computational Efficiency

NS-Pep: De novo Peptide Design with Non-Standard Amino Acids

2025-10-01 · Tao Guo, Junbo Yin, Yu Wang, Xin Gao arxiv

Peptide drugs incorporating non-standard amino acids (NSAAs) offer improved binding affinity and improved pharmacological properties. However, existing peptide design methods are limited to standard amino acids, leaving …

Folding molecular dynamics simulation of T-peptide, a HIV viral entry inhibitor : Structure, dynamics, and comparison with the experimental data

2021-12-28 · Ioanna Gkogka, Nicholas M. Glykos

Peptide T is a synthetic octapeptide fragment, which corresponds to the region 185-192 of the gp120 HIV coat protein and functions as a viral entry inhibitor. In this work, a folding molecular dynamics simulation of pept…