Protein Structure and Dynamics

This study uses molecular dynamics and PCA to analyze coordinated movements in tyrosinase subdomains, showing mutations disrupt stability by disrupting helix coordination.

The paper analyzes a crowdsourced protein design competition, showing participants designed EGFR binders with varying affinities and highlighting challenges in benchmarking and experimental design.

AF2χ predicts protein side-chain rotamer distributions using AlphaFold2, outperforming simulations and existing methods in speed and accuracy.

The paper identifies a new evolutionarily conserved short linear F-actin binding motif and develops a computational pipeline to identify its containing proteins.

HemeFinder predicts heme-binding sites in proteins with high accuracy and speed, demonstrating its utility in identifying heme interactions and pathways.

The paper introduces Foldinsight, a structure-based framework for predicting protein function using structural information, achieving performance comparable to sequence-based models.

MeDCycFold accelerates cyclic peptide structure prediction using a deep learning-based Rosetta distillation model, achieving 49x speedup while maintaining Rosetta-level accuracy.

The paper introduces PepPCBench, a benchmark framework to evaluate AlphaFold3's performance in predicting protein-peptide complexes using a curated dataset.

This paper improves AlphaFold2's ability to predict alternative protein conformations by integrating clustering strategies and evolutionary signals, enabling rational mutation design.

BoltzDesign1 inverts AlphaFold3 to design protein binders for diverse targets without model finetuning, achieving high in silico success rates and structural diversity.

The paper introduces Menger_Curvature, a MDAKit implementation for calculating Menger Curvatures and Local Flexibilities in polymers, showing it is faster than RMSF and effective for analyzing protein dynamics.

The paper benchmarks protein stability predictors using a balanced dataset of AlphaFold structures, showing performance declines when balancing data and highlighting the benefits of molecular dynamics ensembles over static structures.

The paper designs artificial multidomain transmembrane transporters using a combined generative landscape and molecular dynamics approach, achieving native-like functionality.

The paper predicts 3D structures of viral-host protein interactions and identifies evolutionary patterns and functional mechanisms underlying viral pathogenesis.

This paper introduces a coordinate-independent method for encoding protein structures that captures conformational dynamics through dimensionality reduction.

This paper computationally designs multi-component protein filaments that assemble into ordered fibers, validated by cryoEM, and shows regulation via phosphorylation.

This paper introduces a minimal model of protein-protein interactions called Structured Random Binding, showing how nonspecific complexes can evolve into specific ones under certain structural conditions.

This paper develops a generative World Model to simulate protein folding pathways using dihedral rotations and validates it against MD data.

This study identifies pathway-dependent molecular interactions that define distinct physicochemical properties of biomolecular condensates formed by the same IDP.

OpenComplex predicts 3D atomic-level structures of biomolecular complexes using models based on AlphaFold2 and AlphaFold-Multimer, achieving accuracy comparable to or better than existing methods.

This study links molecular properties of LCDs to condensate microstructure and macroscopic material properties using simulations and graph analysis.

This study compares generative AI models for protein design, finding structural models produce more confident structures but less diversity, while language models generate more diverse proteins with lower structural confidence.

This study presents the first structural characterization of NFI-X, revealing its DNA-binding domain structure and key determinants for activity, providing a framework for understanding the NFI family's function and drug design.

The paper uses AlphaFold to analyze protein structures and identifies new lasso motifs, including knotted lasso proteins, and explores their topological complexity and conservation.

This paper evaluates non-coding variants' impact on TF binding using AlphaFold 3 and FoldX, showing structural modeling can provide interpretable insights despite lower accuracy than sequence-based methods.

This paper introduces Martini 3 protein modeling, comparing three structure bias methods for maintaining protein structures in simulations.

The paper evaluates protein structure prediction models on sactipeptides with non-canonical crosslinks, finding limited performance and suggesting the need for physics-informed models.

The paper presents a pipeline combining multi-modal deep learning and AlphaFold3 to improve atomic protein structure modeling from cryo-EM data.

SPfast improves protein structure alignment speed and sensitivity, enabling large-scale comparisons and uncovering new biological insights.

This paper evaluates AlphaFold 3's ability to predict D-peptide chirality, fold, and binding pose in heterochiral complexes and finds it performs poorly, with a high chirality violation rate.