MSA Pairformer predicts protein-protein interface contacts nearly 3 times more accurately than existing methods, despite training only on individual chains.
- Open access
- 2 cites
Moving in Cell.
Rebuilt
MSA Pairformer predicts protein-protein interface contacts nearly 3 times more accurately than existing methods, despite training only on individual chains.
Newest first · last 60 days
ProVADA enables functional protein redesign, converting human renin into an active cytosolic protease and deimmunizing murine antibodies with significant HLA-I binder reduction.
RigidSSL pretraining enhances protein designability by up to 43% and improves conformational ensemble modeling in G protein-coupled receptors.
Optimized power exponents for distance-based variables improve protein-ligand binding affinity prediction, with neural network models reaching R values between 0.800 and 0.900.
Homology-aware cross-validation strategies reveal that current RNA secondary structure prediction methods' apparent superiority is an artifact of splitting strategies.
PlantGeneAnn achieves a new state-of-the-art F1 score of 0.856 at the exon level in plant gene annotation across 20 species, outperforming existing tools.
MassNet is a billion-scale AI-ready mass spectral corpus containing 1.5 billion spectra across 35 species, enabling scalable deep learning in proteomics.
Tangermeme enables extraction of human-interpretable cis-regulatory patterns from deep learning models with high efficiency.
SNACKKSS predicts pharmacologic gene regulators with high accuracy, identifying protein-inhibiting compounds by integrating RNA-Seq and literature data.
Kidzoi, a kidney cell-type-specific S2F model, improves regulatory variant effect prediction by 10-23% over existing models.
Journals this month
Topics moving, week to 23 Oct 2026