Nature Computational Science
Nature Portfolio · Genomics & Bioinformatics · ISSN 2662-8457
- Impact (2-yr)
- 12.7
- h-index
- 91
- Open access
- 47%
- Works
- 1.1K
- Papers this month
- 7
Rebuilt
Latest in Nat Comput Sci
Newest first · last 60 days
AI summary · may contain errors
A geometrically transferable neural-network wavefunction method enables efficient, high-accuracy prediction of ground- and excited-state potential energy surfaces for ethylene, the carbon dimer, methylenimmonium, and a 96-electron rubredoxin active-site model.
AI summary · may contain errors
A physical deep-learning framework on an optoelectronic platform uses state-skipping direct feedback alignment and a single-map 1D dynamics architecture to cut computational resources by ~40% while matching performance.
- Open access
AI summary · may contain errors
Hi-TPH-PLMs fine-tuned on Hi-TPH deliver a 19.2% performance gain over state-of-the-art tools on an external wet-lab test set for T cell receptor–pHLA binding prediction.
AI summary · may contain errors
A structure-coefficient theorem for evolutionary games on dynamic networks with exogenous stochastic transitions shows evolutionary outcomes can be captured by three universal structure coefficients.
- 1 cites
AI summary · may contain errors
HDOCK-Multimer (HDM) substantially outperforms state-of-the-art methods on 3 benchmarks covering 35 large heteromeric complexes, 172 large protein complexes, and 7 CASP15 targets.
- Open access
AI summary · may contain errors
Hessian Interatomic Potentials (HIPs) predict molecular Hessians 1–2 orders of magnitude faster and more accurately than derivative-based approaches, using SE(3)-equivariant deep learning.
AI summary · may contain errors
An active learning–guided workflow for C-H borylation produced a diverse benchmark and enabled symmetry-aware geometric graph neural networks to correctly predict borylation positions in all prospective tests.
- Open access
- 1 cites
AI summary · may contain errors
CellART enables extraction of single-cell information from high-resolution spatial transcriptomics by integrating staining images, spatial data, and single-cell RNA sequencing references across diverse platforms.
AI summary · may contain errors
INSPIRE enables on-chip in situ physical gradient descent for photonic integrated circuits, achieving trained matrices with 0.26% relative error and supporting in situ training beyond native tunable elements.
AI summary · may contain errors
Local uniqueness in modulating complex oscillators generally requires the equal dimension of modulation parameters and targeted spatiotemporal properties, enabling property-focused parameter identification.
AI summary · may contain errors
A theory-grounded physics-transfer learning framework predicts brain morphogenesis and yields reduced-dimensional evolutionary representations, with generalization bounds explaining strong performance in brain feature characterization.
- Open access
AI summary · may contain errors
LLMs show a consistent “geometric blindness” for coordinate information in coordinate-category synthetic datasets, failing to learn it even up to 70 billion parameters and 2 × 10 6 structures.
- Open access
- 13 cites
AI summary · may contain errors
Large language models are used across disciplines as human proxies, and their human-likeness is not one property because each role demands distinct validity criteria.
AI summary · may contain errors
MutexaGPT, an open-access multi-agent LLM platform, translates plain-English enzyme engineering intuition into physics-based simulations and actionable variant designs for two protein engineering tasks.
AI summary · may contain errors
CrysVCD, a valence-constrained diffusion framework for crystal generation, attains 85% metastability (E hull < 0.1 eV per atom) and 68% phonon stability after stability fine-tuning.
AI summary · may contain errors
Self-supervised pretraining on equilibrium conformers substantially improves generative transition-state prediction for previously unseen transition metal complexes, lowering median TS-geometry RMSD on Transition1x-TMC reactions.
AI summary · may contain errors
Integrating gray and white matter iEEG signals enables high-performance, cross-task decoding of continuous valence and arousal in 18 individuals, with robust real-time decoding in four new individuals.
Browse JClub
Journals this month
- bioRxiv6386
- medRxiv1756
- PNAS556
- JAMA352
- Sci Adv330
- Cell Rep314
- Environ Res213
- BMJ209
- EHJ206
- Nat Commun172
- NAR172
- Plant Cell Environ167
- Environ Pollution153
- Science152
- NEJM148
- New Phytol145
- Curr Biol130
- Nature127
Topics moving, week to 17 Nov 2026
- Single-cell and spatial transcriptomics10
- Gut microbiota and health6
- CAR-T cell therapy research6
- Pancreatic and Hepatic Oncology Research5
- Genomics and Chromatin Dynamics5
- Inflammatory Bowel Disease5
- Lung Cancer Treatments and Mutations5
- Antibiotic Resistance in Bacteria5
- Genomics and Phylogenetic Studies4
- Lipoproteins and Cardiovascular Health4
By field
- General Science & Medicine22
- Clinical Medicine25
- Cardiology & Vascular Biology12
- Endocrinology & Metabolism10
- Gastroenterology & Hepatology7
- Public & Global Health4
- Cell & Molecular Biology30
- Immunology & Inflammation14
- Neuroscience14
- Psychiatry & Mental Health8
- Genomics & Bioinformatics29
- Pharmacology & Drug Discovery5
- Microbiology & Mycology34
- Oncology22
- Ecology & Evolution5
- Protected Cropping & Horticultural Science5
- Plant Physiology & Photobiology19
- Student Research1
