Protein Structure and Dynamics

This paper uses writhe to analyze conformational ensembles and dynamics of intrinsically disordered proteins, showing it improves Markov state models and enables equivariant neural network sampling.

aSAM, a latent diffusion model trained on molecular dynamics, generates heavy-atom protein ensembles; its temperature-conditioned variant aSAMt reproduces mdCATH thermal unfolding and extrapolates beyond training temperatures.

The paper improves AlphaFold2 and AlphaFold3 predictions using MSA engineering, extensive model sampling, and ensemble QA methods, achieving high accuracy in CASP16.

DynaRepo provides a repository of macromolecular conformational dynamics with simulations and analyses for over 450 complexes and 270 proteins.

AlphaSync provides updated AlphaFold structure models for 39,961 UniProt proteins and enhances 2.6 million structures with residue-level features.

This study identifies molecular and evolutionary determinants for ALDODOM protein interactions, showing that a C-terminal helix and conserved residues enable Mud binding in Drosophila Adducin.

This study improves stereochemical accuracy in protein-ligand complex structure prediction by addressing limitations in AlphaFold3 and Boltz-1 through a restraint-guided inference method.

This paper evaluates blind predictions of dynamic water and ion ensembles around RNA using cryo-EM densities from CASP16, finding that molecular dynamics approaches perform best.

An atom-level graph neural network that denoises corrupted structures generates small-molecule and protein-ligand conformational ensembles, and its active-site pre-organization metric enriched for more active designed retroaldolases.

This paper improves the Gaussian Network Model by integrating local and allosteric factors to better predict residue-residue correlations and B-factors in proteins.

This study compares intrinsic dynamics of PAS domains in bHLH-PAS transcription factor complexes using ENMs and MD simulations, revealing conserved self-coupled motions and context-specific directly coupled motions.

This paper evaluates progress in predicting protein complex structures in CASP16, finding moderate improvements but ongoing challenges in model ranking, stoichiometry prediction, and handling difficult targets.

MDZip compresses molecular dynamics trajectories using neural networks, achieving high compression with minimal loss of structural and ensemble-level information.

SiteAF3 improves AlphaFold3's accuracy for site-specific folding of biomolecular complexes using conditional diffusion, outperforming AlphaFold3 on various datasets.

This paper evaluates methods for modeling multi-state protein and RNA systems in CASP16, finding improved accuracy for some targets but persistent challenges in capturing structural details for others.

This paper introduces SST2, a new enhanced sampling method for molecular dynamics simulations that improves efficiency for large biomolecular systems by selectively scaling interactions within and around the molecule.

The paper analyzes how phosphorylation affects protein structure and dynamics using structural data from the PDB, finding that it often induces stabilizing conformational changes and modulates residue fluctuations.

This paper introduces a Monte Carlo framework to simulate biomolecular condensate nucleation in the grand canonical ensemble, revealing classical and nonclassical nucleation mechanisms in FUS-LC and NDDX4 proteins.

This paper develops a method to compute residue-scale elastic pseudoenergy in proteins and applies it to RuvB to elucidate the energetic basis of its hand-over-hand mechanism.

This study investigates calcium-induced Sorcin aggregation using stopped-flow light scattering, finding that aggregation is reversible, cooperative, and concentration-dependent.

The paper integrates bioinformatics, molecular dynamics, and single-molecule FRET to study ligand-dependent conformational dynamics of membrane proteins.

This paper introduces a method to guide protein engineering using machine learning that leverages dynamics-property relationships from molecular dynamics simulations with minimal experimental data.

The paper uses molecular dynamics simulations to study how mutations in the bacteriophage λ Cro repressor alter its DNA-binding properties and regulatory function.

AQuaRef uses machine learning to refine protein structures with quantum accuracy at lower computational cost, improving geometric quality and experimental data fit.

This paper introduces PIGEON-FEATHER, a method to calculate free energies of protein conformational ensembles from HX/MS data, revealing evolutionary differences in enzyme inhibitors and ligand-induced reweighting in a protein-DNA complex.

The paper predicts the structural flexibility of antibody and T-cell receptor CDR3 loops using a new method trained on a large dataset of conformations.

The paper evaluates computational methods for predicting interdomain pose distributions in D-L-D proteins using NMR and SAXS data, finding none accurately match experimental results.

PPI-ID maps interaction domains and motifs onto protein structures to predict protein-protein interactions with high accuracy using AlphaFold-Multimer.

This paper introduces a hybrid ML framework to predict conformational properties of IDPs using sequence and physical features, achieving improved accuracy and interpretability.

The paper shows that multivalent interactions among disordered regions can explain condensate specificity, using a polymer physics model and experimental validation.