Guardado en:
Detalles Bibliográficos
Autores principales: Nori, Divya, Parsan, Anisha, Uhler, Caroline, Jin, Wengong
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:https://arxiv.org/abs/2505.21241
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866913861401051136
author Nori, Divya
Parsan, Anisha
Uhler, Caroline
Jin, Wengong
author_facet Nori, Divya
Parsan, Anisha
Uhler, Caroline
Jin, Wengong
contents Protein binder design has been transformed by hallucination-based methods that optimize structure prediction confidence metrics, such as the interface predicted TM-score (ipTM), via backpropagation. However, these metrics do not reflect the statistical likelihood of a binder-target complex under the learned distribution and yield sparse gradients for optimization. In this work, we propose a method to extract such likelihoods from structure predictors by reinterpreting their confidence outputs as an energy-based model (EBM). By leveraging the Joint Energy-based Modeling (JEM) framework, we introduce pTMEnergy, a statistical energy function derived from predicted inter-residue error distributions. We incorporate pTMEnergy into BindEnergyCraft (BECraft), a design pipeline that maintains the same optimization framework as BindCraft but replaces ipTM with our energy-based objective. BECraft outperforms BindCraft, RFDiffusion, and ESM3 across multiple challenging targets, achieving higher in silico binder success rates while reducing structural clashes. Furthermore, pTMEnergy establishes a new state-of-the-art in structure-based virtual screening tasks for miniprotein and RNA aptamer binders.
format Preprint
id arxiv_https___arxiv_org_abs_2505_21241
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle BindEnergyCraft: Casting Protein Structure Predictors as Energy-Based Models for Binder Design
Nori, Divya
Parsan, Anisha
Uhler, Caroline
Jin, Wengong
Machine Learning
Protein binder design has been transformed by hallucination-based methods that optimize structure prediction confidence metrics, such as the interface predicted TM-score (ipTM), via backpropagation. However, these metrics do not reflect the statistical likelihood of a binder-target complex under the learned distribution and yield sparse gradients for optimization. In this work, we propose a method to extract such likelihoods from structure predictors by reinterpreting their confidence outputs as an energy-based model (EBM). By leveraging the Joint Energy-based Modeling (JEM) framework, we introduce pTMEnergy, a statistical energy function derived from predicted inter-residue error distributions. We incorporate pTMEnergy into BindEnergyCraft (BECraft), a design pipeline that maintains the same optimization framework as BindCraft but replaces ipTM with our energy-based objective. BECraft outperforms BindCraft, RFDiffusion, and ESM3 across multiple challenging targets, achieving higher in silico binder success rates while reducing structural clashes. Furthermore, pTMEnergy establishes a new state-of-the-art in structure-based virtual screening tasks for miniprotein and RNA aptamer binders.
title BindEnergyCraft: Casting Protein Structure Predictors as Energy-Based Models for Binder Design
topic Machine Learning
url https://arxiv.org/abs/2505.21241