protein sequence

This paper develops an anaerobic cell-free protein synthesis platform to study oxygen-sensitive proteins, demonstrating its ability to produce functional enzymes under anoxic conditions.

This paper designs de novo ligases that achieve site-specific ubiquitylation of a target protein with high selectivity and efficiency.

The paper develops a method combining evolutionary and protein language model representations to predict protein stability changes with thermodynamic consistency, improving accuracy and physical consistency.

The paper develops a method to predict phenotype modulation by analyzing amino acid clustering and epistasis in protein sequences, showing that these factors improve prediction accuracy.

PPI-Caliper controls false discovery rate in protein interaction prediction using conformal calibration without retraining models.

The paper decodes protein N-glycosylation sites from protein language model embeddings using a zero-shot framework, revealing biological insights and disease associations.

This paper introduces a direct enzymatic bioconjugation platform using the flavin transferase ApbE, enabling site-specific conjugation of diverse payloads without intermediate chemical steps.

This paper fine-tunes a protein language model on nanobody thermostability data and uses sparse autoencoders to uncover biophysical rules governing stability, identifying both known and novel stabilizing features.

The P450 ApyO crosslinks a 10-residue C-terminal peptide, and substitutions in the YLY motif switch the linkage among C−C, N−C and C−O bonds.

This paper redesigns PET hydrolase PHL7 using MPNN models to enhance its catalytic activity at lower temperatures, achieving efficiency comparable to its activity at higher temperatures.

The paper predicts residue-level protein-protein interactions between hepatitis B virus core and envelope proteins using deep learning and coevolution signatures, comparing results with experimental data.

This study shows disease-linked mutations in hnRNPA2B1 accelerate condensate maturation and promote amyloid aggregation by altering phase separation and fibril formation.

This paper introduces an interpretable machine learning method to derive optimized collective variables for protein-ligand binding-unbinding transitions, achieving accurate free energy estimates and residence times with low computational cost.

The paper explains how autotransporter proteins avoid a kinetic trap during secretion by using a computational framework to reveal hidden conformational states.

The paper introduces MEM-CALVADOS, a residue-level model for flexible proteins at membrane interfaces, validated against experimental data and applied to signaling protein assemblies.

This study identifies a cleavable signal peptide in TMEM165 essential for its Golgi targeting and membrane topology, showing that N-terminal deletions restore function and suggest the signal peptide's role in biogenesis.

Ancestral sequence reconstruction of a bacterial decarboxylase gave enzymes with up to 294-fold longer half-life at 45 °C, higher activity, and 99.7% ee on a vinyl malonate substrate.

This paper demonstrates an AI system that autonomously interacts with biological systems to discover new protein functions through continuous experimental feedback.

The paper uses deep learning to show estradiol binds androgen receptors more strongly than testosterone, revealing evolutionary conservation and potential therapeutic implications.

This study shows that mitochondrial presequences function in cytosolic quality control and trigger a heat shock response when mitochondrial import is inhibited.

The paper uses computational methods to redesign β-lactoglobulin, improving its stability and BCAA content, with experimental validation of the results.

Trained on missing NMR amide assignments in the Biological Magnetic Resonance Data Bank, Dyna-1 (ESM-3 layer 22 embeddings) predicts micro-to-millisecond conformational exchange seen in relaxation data.

The paper shows that mimicking oxidative modifications at vimentin residues C328 and S49 induces phase separation, revealing a role for glycosylation and phosphorylation in regulating this process.

This study identifies the flexible N-terminal domain of Sorcin as a critical regulator that enhances Ca2+ responsiveness by stabilizing its active conformation through dynamic contacts.

The paper uses genome-informed machine learning to discover a new enzyme, PTC, that catalyzes phenazine thioconjugation from sparse data.

This study identifies surface secondary binding sites in β-glucosidase BglB that regulate glucose inhibition and catalytic activity through long-range structural and energetic changes.

The paper compares the distribution of young de novo proteins, old canonical proteins, and randomized sequences in sequence space using k-mer-based distances, finding distinct patterns and potential pathways between classes.

This study develops a machine learning model to predict protein stability from N-terminal sequences and identifies stability-determining motifs in annotated and alternative isoforms.

This paper introduces Sprox-seq, a multi-omic technique that maps protein interactions and mRNA in human germinal centers, revealing distinct B cell maturation pathways and spatially-enriched immune regulatory complexes.