Yuguang Mu

Filling the Missing Puzzle: Integrating Interaction and Evolutionary Information to Enhance Protein Fitness Prediction in Inverse Folding

This paper introduces zPolaron, a graph neural network for functional protein design that integrates structural, interaction, and evolutionary information, improving fitness prediction beyond structural compatibility.

CrossAffinity: A Sequence-Based Protein-Protein Binding Affinity Prediction Tool Using Cross-Attention Mechanism

CrossAffinity predicts protein-protein binding affinity using a sequence-based cross-attention model, outperforming both structure-based and sequence-based methods on multiple test sets.

Metadiffusion: inference-time meta-energy biasing of biomolecular diffusion models

Metadiffusion introduces a meta-energy biasing layer to diffusion models to generate diverse biomolecular conformations without retraining, matching molecular dynamics simulations.

Benchmarking AlphaFold3-like Methods for Protein-Peptide Complex Prediction

This study benchmarks AlphaFold3-like methods for predicting protein-peptide complex structures, showing improved accuracy but challenges in binding site prediction.

Systematic benchmarking of deep-learning methods for tertiary RNA structure prediction

This study benchmarks deep learning methods for RNA tertiary structure prediction, finding that ML methods outperform non-ML methods but struggle with novel RNAs and non-Watson-Crick pairs.

Structural insights into physiological activation and antagonism of melanin-concentrating hormone receptor MCHR1

This study presents cryo-EM structures of MCHR1 in active and inactive states, revealing ligand recognition and activation mechanisms to aid drug development.

ProAffinity-GNN: A Novel Approach to Structure-based Protein-Protein Binding Affinity Prediction via a Curated Dataset and Graph Neural Networks

This paper introduces ProAffinity-GNN, a deep learning framework for predicting protein-protein binding affinities using graph neural networks and a curated dataset of protein complexes with experimentally determined affinities.

Benchmarking Reverse Docking through AlphaFold2 Human Proteome

This study benchmarks reverse docking pipelines using AlphaFold2 human proteome to predict drug-target interactions and off-target effects, finding that Glide_sfct (PS) performs best.

Protein language models are performant in structure-free virtual screening

This paper shows that protein language models can achieve screening performance similar to structure-based methods without requiring 3D protein structures.

Quantitative intra-Golgi transport and organization data suggest the stable compartment nature of the Golgi

The paper challenges the classical cisternal progression model by showing that intra-Golgi transport velocity varies and Golgi residence times differ for different cargos, suggesting a stable compartment model.

A New Paradigm for Applying Deep Learning to Protein-Ligand Interaction Prediction

This paper introduces IGModel, a deep learning framework that predicts protein-ligand docking RMSD and binding strength, achieving state-of-the-art performance on multiple benchmarks.

Variational graph encoders: a surprisingly effective generalist algorithm for holistic computer-aided drug design

The paper introduces a generalist model for drug design that predicts multiple properties simultaneously, achieving faster virtual screening and drug development with explainability.

In silico repurposed drugs against monkeypox virus

This study identifies FDA-approved drugs that may repurpose to treat monkeypox by predicting their binding to viral proteins critical for replication.