Protein Structure and Dynamics

This study uses molecular dynamics simulations to show that VISTA's conformational switching between Up and Down states regulates its immune checkpoint function through CC' loop accessibility.

This paper investigates the temporal dynamics of discrete protein diffusion models, revealing a decoupling between structural alignment and property optimization, and identifies a structural collapse regime due to Coulombic repulsion.

This study investigates the conformational flexibility of staphylokinase variants to explain their reduced immunogenicity and altered properties.

This study evaluates seven single-point mutants of β-glucosidase B using computational design and experimental analysis, finding weak correlations between predicted thermodynamic stability and enzyme properties.

The paper introduces FELS, a method to screen free energy landscapes for disordered proteins, and applies it to tropoelastin, revealing its conformational ensemble and resolving conflicting models.

The paper introduces AtomPaint, a full-atom diffusion model for designing protein/non-protein binding interactions, showing it outperforms random guessing in sequence recovery.

The paper introduces MARS, a simulation framework that resolves conformational heterogeneity in IDPs by integrating experimental data from SAXS and PRE, revealing two major states in ER-NTD.

The paper shows that peptides made from non-standard amino acids can form protein-like secondary structures, expanding the known sequence-structure space beyond Earth's 20 amino acids.

The paper introduces an AI framework for protein interaction design using KAN-based PU learning, achieving improved stability and validation metrics compared to traditional methods.

This study shows a short peptide can replace an essential enzyme in yeast, demonstrating peptide-based catalysis as a possible precursor to modern protein enzymes.

This paper identifies amphiphilic motifs that encode both fluid lyotropic phases and amyloid formation, linking functional condensation with pathological aggregation.

UdonPred predicts per-residue protein disorder from sequence using embeddings from ProstT5, showing that annotation quality, not model capacity, drives performance in continuous disorder prediction.

LocAlign uses geometric deep learning to align protein structures and identify functional motifs, achieving high accuracy for both similar and dissimilar folds.

The paper updates the MAD web server with new tools and an enhanced database for building and simulating coarse-grained polymer models using the Martini force field.

The paper introduces CoEVFold, a pipeline that identifies and visualizes protein coevolution using MSA data and provides tools for analyzing coevolution in proteins and complexes.

Surf2Spot predicts protein-protein and nanobody-antigen interaction hotspots using a deep learning model that integrates sequence, structure, and surface properties, improving performance over existing methods.

ESMDisPred predicts intrinsically disordered proteins by integrating sequence embeddings with structural data, achieving high accuracy on benchmark datasets.

FrustraMPNN rapidly predicts single-residue frustration in proteins at proteome scale with high accuracy, enabling new insights into protein function and dynamics.

This paper designs high-affinity protein binders using a computational framework combining flow matching and in silico maturation, achieving picomolar and nanomolar affinities without experimental screening.

This paper uses atomistic simulations to show that ATP hydrolysis in myosin II stores mechanical energy in functional loops through hydrogen bond rearrangements, independent of heat release.

This study uses Constraint Network Analysis to compare the mechanical rigidity of wild-type EGFR in apo and gefitinib-bound states, finding that ligand binding causes localized rigidification without large-scale structural reorganization.

The paper evaluates the neutrality of amino acid homorepeats in yeast, finding hydrophobic repeats are harmful while hydrophilic repeats suppress proteotoxic stress and define compatible sequence space.

This paper introduces a GAN-based method to generate molecular dynamics trajectories with kinetic fidelity, enabling efficient sampling of rare events in complex systems.

This paper introduces ProtProfileMD, a model that predicts protein dynamics from sequences, improving homology detection by capturing conformational changes through 3Di tokens.

This study investigates how protein folding landscapes change within biomolecular condensates using atomistic simulations and Bayesian optimization models, revealing sequence-dependent folding behavior influenced by multivalent interactions and crowding.

AFsample3 improves protein conformational sampling by generating and selecting multiple states, outperforming AlphaFold3 and AFsample2 in accuracy and diversity.

PathDiffusion models protein folding pathways using evolution-guided diffusion, accurately reconstructing folding events and demonstrating versatility across multiple applications.

This paper introduces ProStab-Former, a deep learning model for protein stability prediction that improves upon existing methods by integrating sequence and structure information, achieving competitive performance and efficient mutation prediction.

This paper introduces IDPFold2, a framework for predicting conformational ensembles of proteins, including multidomain proteins and complexes, and shows it outperforms existing methods.