protein sequence

The paper identifies a refined sticker-spacer grammar in phase-separating IDRs that correlates with pathogenic missense variants and explains mutational vulnerability in disordered proteins.

The paper uses AI to compare biological information from UniProt records and their cited literature for human kinases, finding literature provides better coverage for certain topics.

The paper designs proteins using reduced amino acid alphabets and shows that 8 amino acids can form stable proteins, while smaller alphabets fail in experimental validation.

The paper analyzes insect odorant receptor (OR) diversity across 115 species using sequence, structure, and docking profiles, finding associations with diet and habitat and identifying a conserved Orco domain.

The paper surveys TOR and AMPK across 210 eukaryotes, revealing lineage-specific variations in component repertoires, sequence features, and component associations.

OmegaTherm predicts mutation-induced protein stability responses by jointly modeling folding free energy and melting temperature, achieving state-of-the-art performance in prediction and variant ranking.

MuseDrift generates protein variants with specified sequence identity to a wild-type sequence, achieving higher structural prediction confidence and better mutation-effect scores than existing methods.

The paper develops a quantitative statistical mechanical model to predict protein interactions from sequences and uses it to analyze the structure and organization of human signaling networks.

OpenKinetics Predictor unifies thirteen enzyme kinetic predictors behind one interface and API; on E. coli iML1515, eight kcat predictors correlate pairwise at r 0.08 to 0.67.

ProDive detects cross-family local conservation using a pHMM-based framework, identifying 318,289 high confidence segment correspondences across 25,545 Pfam families.

Proteome-wide variant calling builds amino acid phylogenomic alignments orders of magnitude faster than a BUSCO marker pipeline, with similar tree signal in three of four test taxa.

BRIDGE-AD integrates multimodal AD datasets to prioritize disease effectors, identifying 19 functional clusters and validating SCARB2 as a novel candidate.

LoGoPPI infers protein-protein interactions from sequence using a combination of global and local representations, achieving faster inference than state-of-the-art models.

This paper introduces DEFT, a method that combines sequence- and structure-based approaches to improve EC number prediction accuracy and enable high-throughput enzyme repertoire annotation.

MetExPred predicts drug metabolism and excretion by integrating molecular and protein representations, achieving strong performance across multiple endpoints.

This paper introduces AB-NW, an adaptive-banding Needleman-Wunsch alignment method that improves speed while maintaining accuracy on challenging sequence alignment tasks.

The paper analyzes the correlation between codon degeneracy and amino acid frequencies across genetic codes, revealing consistent patterns of over- and underrepresentation of amino acids across domains of life.

This paper introduces an ultrasensitive luciferase reporter to better detect low-level mRNA expression of lipid nanoparticles in non-hepatic tissues.

This paper introduces a PhyloPotts model that combines Potts models with phylogenetics to improve ancestral sequence reconstruction, though it does not consistently reduce bias in phylogenetic estimates.

AnewDDE integrates AI tools for biomolecular interaction modelling and drug design, demonstrating improved performance in structure prediction, binder identification, affinity estimation, and ADMET prediction.

COCOA-Tree reanalyzes coevolution datasets to uncover evolutionary properties of coevolving residues and shows different metrics identify distinct groups.

This study investigates how sequence design parameters influence phase separation and mixing behavior of intrinsically disordered proteins (IDPs) when mixed.

The paper investigates the evolutionary origin of NELL-Robo interactions and shows that dual-ligand responsiveness is conserved in chordates but not protostomes.

Minerva extracts RNA base-pairing, protein-contact and repeat maps from a genome language model without alignments, flagging unannotated structured loci and a phage ncRNA array reverse-transcribed into short cDNAs.

NeoToxPred improves toxicity classification by using a protein language model with carefully curated negative controls, outperforming existing methods on both internal and external benchmarks.

This paper argues for evaluating protein language models through prospective experimental success rather than retrospective predictive accuracy.

The paper introduces Cerebra-Epistasis, a framework that improves multi-mutant fitness prediction by modeling epistatic effects and integrating structure-aware sequence analysis.

The paper introduces a lossless compression algorithm for protein databases, improving memory efficiency and classification accuracy in Centrifuger for metagenomic analysis.

The paper introduces CAMP-14-3-3, a framework that improves 14-3-3 binding prediction by integrating multi-scale biological features and protein language model embeddings, outperforming existing methods.

MutCleaner standardizes mutation datasets for variant effect prediction by cleaning, validating, and organizing mutation records across 16 datasets with over 12.14 million entries.