Protein Structure and Dynamics

This paper designs de novo proteins that undergo small-molecule-induced conformational changes, demonstrating strong binding affinity and specificity for exatecan.

This paper introduces seed-guided diffusion-based design to improve structural diversity and target contact in antitoxin protein binders, achieving functional binders with enhanced selectivity.

This paper introduces CLSS, a contrastive learning model that co-embeds protein sequences and structures, revealing evolutionary relationships and structural transitions.

This paper introduces TEA, a temperature-dependent coarse-grained model for simulating thermoresponsive behavior of disordered proteins, showing it accurately captures LCST and UCST transitions.

TM-Vec 2s improves protein remote homology detection by being faster and more accurate than previous methods, enabling efficient large-scale database searches.

This study investigates how intra-protein interfaces influence folding dynamics and mechanical stability in a designed repeat protein using single-molecule techniques and simulations.

This study benchmarks deep learning models' ability to predict mutation-induced fold switching in proteins using an NMR-characterized dataset of fold-switching mutants.

The paper identifies a buried charge-network GFP-fold family present in prokaryotes and eukaryotes, revealing convergent evolution of a conserved residue network.

This paper introduces Boltz2ESI, a framework predicting enzyme-substrate interactions using a foundation model, outperforming existing methods by capturing active-site plasticity and evolutionary context.

The paper introduces an alphabet of allostery using transfer entropy to predict allosteric-site character and conformational rewiring across proteins.

SPINDLE uses deep learning to predict protein dynamics from single-field NMR data, outperforming traditional methods and reproducing experimental results for 50 proteins.

TriFlow, discrete flow-matching sequence designer with global three-track structure attention, raises in silico binder success over ProteinMPNN in RFdiffusion and BindCraft-style pipelines on 500 targets and flags functional sites.

The paper identifies proteins that undergo significant isotropic volume expansion or contraction using intramolecular distance scoring analysis and shows their role in ligand binding and protein folding.

This paper introduces pHoptNN, a graph neural network predicting enzyme pH opt from structure, achieving lower error than sequence-based methods.

This study evaluates protein language model embeddings for estimating structural similarity in the protein-sequence twilight zone and finds that ProstT5 performs best.

This study benchmarks AlphaFold3's antibody-antigen binding predictions using experimental data and identifies biases and false positives in its performance.

The paper introduces Reseek, a protein structure search algorithm achieving the highest accuracy with practical scalability for large databases.

This paper introduces EffieDes, a neuro-symbolic AI framework for protein sequence design that combines deep learning with automated reasoning to generate sequences with complex functional properties.

This paper introduces ILMC, a sampling strategy for protein language models that improves diversity while steering generation toward desired properties.

This paper introduces SPARKfold, a microfluidic platform to measure protein unfolding rates for thousands of variants, validating its accuracy and demonstrating its utility in studying kinetic stability.

This paper uses molecular dynamics simulations of AAA+ unfoldases to show structural asymmetry and allosteric networks generate kinetic asymmetry, enabling directional substrate translocation despite stochastic ATP hydrolysis.

This paper presents a computational framework to design catalytic antibodies by transferring enzyme active site geometry into antibody CDRs, enabling controlled catalytic residue arrangement and improved chemical turnover.

The paper introduces Pinc, a simple probabilistic score to assess interface confidence in AlphaFold models, showing it matches or exceeds complex methods' performance.

This paper introduces the Ordered Gromov-Hausdorff (OGH) metric to compare protein structures by incorporating residue order, showing it outperforms TM-score in detecting homologous proteins.

The paper uses Hodge decomposition to analyze information flow in KRAS proteins, showing that mutations reposition information sources while maintaining a hierarchical flow structure.

Carbonara uses SAXS data to predict alternative protein conformations, enabling more accurate MD simulations of solution-state dynamics.

This paper evaluates limitations of deep learning in biomolecular structure prediction, highlighting the benefits of explicit ion modeling for improving conformational accuracy.

This study investigates the formation of the earliest structure during protein folding on the ribosome and identifies sub-ensembles that may facilitate co-translational folding.

This study benchmarks metal-parameterization strategies for simulating the PAN endonuclease domain to understand resistance mutations in influenza virus.

This study develops a desolvation-aware CG model to investigate LLPS of intrinsically disordered proteins, revealing how desolvation affects phase behavior and kinetics.