AI summary · may contain errors
NCFlow is a flow-matching generative model that can place arbitrary noncanonical amino acids into a protein backbone using only atom types and bond connectivity, generalizing beyond training.
- 14 cites
Moving in Bioinformatics, bioRxiv and PLOS Computational Biology.
Rebuilt
AI summary · may contain errors
NCFlow is a flow-matching generative model that can place arbitrary noncanonical amino acids into a protein backbone using only atom types and bond connectivity, generalizing beyond training.
AI summary · may contain errors
Archaeal proteomes contain 987 of 2,457 ECOD X-groups, capturing roughly 40% of known fold diversity within a single domain of life across 65 archaeal classes.
AI summary · may contain errors
AtomLM-9b achieves 75.01% success on the FoldBench protein–protein task, matching AlphaFold 3’s reported 72.93% using an all-atom structural tokenizer interface.
Newest first · last 60 days
AI summary · may contain errors
Germinal Multi designs and rescues antibodies that jointly optimize one paratope for multiple antigenic variants, yielding breadth against resistance-associated EGFR variants and distinct historical SARS-CoV-2 variants.
AI summary · may contain errors
Multiconformer ensemble models built from high-resolution X-ray crystallography predict experimental binding entropies for over 70 protein–ligand pairs across 12 proteins, with strong linear correlation.
AI summary · may contain errors
Pallatom2, an atom-to-atom generative binder model, identifies 112 active designs among 210 candidates (53.3%) and yields binders for all ten experimentally evaluated targets.
AI summary · may contain errors
R60P in the apo FimH lectin domain shows the strongest destabilization and the only reproduced local structural change, disrupting the beta-strand at residues 60–61 and three hydrogen bonds.
AI summary · may contain errors
In the substrate-free, fully copper-loaded laccase from Trametes versicolor, two active-site loops (L1 and L2) form metastable states that interconvert sub-millisecond, with T114A destabilizing five states by 3.2–8.2 kcal/mol.
AI summary · may contain errors
Tycholyzer streamlines analysis of NanoTemper Tycho NT.6 nanoDSF data by combining multiple runs and extracting apparent protein melting temperatures using fitting or derivative extrema.
AI summary · may contain errors
RFD4-Proteina improves in silico protein design success rates at higher generation throughput than all previous evaluated methods, using intrinsic and extrinsic reasoning in a shared chemical–conformational latent space.
AI summary · may contain errors
Seven structural classes (A–G) of prokaryotic DNA methyltransferases were identified by catalytic-domain 3D topology, using 38 experimentally determined structures plus AlphaFold models.
AI summary · may contain errors
Membrane-bound Tumor Necrosis Factor (TNF) stabilizes the symmetric extracellular-domain state more than soluble TNF, with lipid engagement restricting conformational heterogeneity.
AI summary · may contain errors
SLIP (Synthetic Landscape Inference for Proteins) constructs tunable synthetic protein fitness landscapes using Potts models, producing difficulty levels suitable for benchmarking ML-guided directed evolution pipelines.
AI summary · may contain errors
About 14% of predicted RfaH homologs are predominantly in an active, NusG-like fold state, and in vivo they show constitutive activity comparable to known RfaH mutants.
AI summary · may contain errors
ABHD5 uses a sequential dual-site recognition of lipid droplet and endoplasmic reticulum membranes, with an N-terminal anchor and a lid-helix contact driving a conformational switch.
AI summary · may contain errors
A conserved short linear motif in the cytosolic Tom22 domain forms a transient -helical PPDE that displaces preproteins above the TOM pore to drive outer-membrane import.
AI summary · may contain errors
BioLM-Score improves DEKOIS 2.0 screening endpoints over a matched geometry-only baseline, but affinity scoring and ranking do not improve uniformly under the same controlled comparisons.
AI summary · may contain errors
HieraPTM-HieraFuncPhos predicts functional phosphosites with biophysically grounded mechanical patterns at HSP90AA1 T36 and T115, matching published NMR observations.
AI summary · may contain errors
N-terminal intrinsically disordered regions suppress cotranslational protein degradation, with the human hRpl7A IDR reducing CFTR CTPD by over 80% when transplanted.
AI summary · may contain errors
ESM2 embeddings capture the sequence grammar of binary patterning in four-helix bundle designs, revealing a jellyfish-shaped manifold across 128 heptad patterns.
AI summary · may contain errors
DynamicsPLM, a 35-million-parameter dynamics-enriched protein language model, predicts zero-shot mutation effects from sequence by balancing evolutionary plausibility with structural dynamics disruption.
AI summary · may contain errors
Posaconazole binds two distinct sites on Parechovirus-A3 capsids—VP0–VP1 in US WI-09 and VP3–VP1 in 9C-CMKC and US WI-09 POSA P14—shifting VP0’s contribution from 86% to 0%.
AI summary · may contain errors
CENP-A centromeric nucleosomes show a more stable histone core and reduced histone-DNA contacts than canonical H3 nucleosomes, with distinct tail engagement dynamics.
AI summary · may contain errors
HandShake, a sequence-only partner-specific protein-protein interface predictor, achieves AUROC 0.811 and MCC 0.367 at 30% redundancy, exceeding the best prior sequence-only benchmark.
AI summary · may contain errors
Two-dimensional replica-exchange simulations of Abl kinase with Abltide reveal five encounter regions and six intermediate states linking initial contact to the bound pose.
AI summary · may contain errors
An unsupervised autoencoder-based deep learning pipeline classifies cryo-EM/ET particle images by jointly optimizing pose and structure, separating real particles from contamination and protein classes.
AI summary · may contain errors
Natural protein structures show a narrow mutational profile with high mutational robustness but limited structural variation, unlike RNA, when secondary-structure distributions are compared under point mutations.
AI summary · may contain errors
A shared latent-space deep learning model redesigned PDZ-binding peptides for DLG4-PDZ3, yielding 100 top candidates with 84 in the top 1% of measured interaction fitness.
AI summary · may contain errors
SLAE-ERR predicts its own backbone 15N chemical-shift error with a mean absolute error of 0.76 ppm, using embeddings from SLAE trained on NMR-linked dynamic regions.
AI summary · may contain errors
ARIES builds PLM-based multiple sequence alignments that achieve higher accuracy than state-of-the-art methods, especially in low-identity regimes, while scaling almost linearly with sequence count.
AI summary · may contain errors
A second-generation CUDA PME-CpHMD implementation in Amber increases throughput 1.5–2× across protein systems of 45–1,338 residues while keeping CpHMD overhead to 12–18% versus conventional MD.
AI summary · may contain errors
In Bacillus subtilis, CdaR’s extracellular YbbR domains undergo a reversible conformational switch between pH 4.5 and 5.5, centered on a hinge between domains 1 and 2.
AI summary · may contain errors
Evolutionarily Conserved Fragments (ECFs) built from 5,396 RefSeq bacteriophage genomes annotate 50% of phage representative proteins, expanding functional coverage beyond ECOD and Pfam.
Journals this month
Topics moving, week to 17 Nov 2026
By field