Top 40
View all ›BACKFILLED Code and industry signals are partial for backfilled days.
Rank 1
Epigenomics
Sequence-to-function models decode regulatory rules and 3D chromatin organization
Sequence-based models are now paired with rule-level interpretation and Hi-C, Micro-C, and imaging-based 3D genome inference.
3 preprints · 3 published · ↑ 2.3× vs 4-wk avg
Rank 2 ML / AI methods
Learned protein and peptide representations predict binding, stability, and kinetics
Transfer learning from pretrained protein encoders is extending deep predictors to peptides, GPCR mutational effects, and out-of-distribution enzyme kinetics.
6 preprints · 1 published · ↑ 2.2× vs 4-wk avg
Rank 3 Genomics
Haplotype-based simulation and reference tools drive population variant inference
Recent preprints and publications pair real-haplotype simulation with mitochondrial and ancestry tools built on public reference resources.
4 preprints · 1 published · ↑ 2.1× vs 4-wk avg
Rank 4 ML / AI methods
Diffusion denoising reframed as transport map optimization for 3D molecule generation
Recent preprints pair generative denoising, Bayesian network inference, and geometry-aware surrogates, each evaluated on synthetic or cross-subject data.
3 preprints · 1 published · ↑ 1.9× vs 4-wk avg
Rank 5 Tools & infrastructure
Curated metabolic and nutritional resources enable computational microbial inference
Reaction coverage, strain embeddings, and cobamide distributions are now curated inputs for microbial metabolic and ecological inference.
2 preprints · 2 published · ↑ 1.8× vs 4-wk avg
Rank 6 Supervised deep models predict RNA modifications, isoforms and regulatory links↑ 1.7×ML / AI methodsRank 7 Benchmarks and long-read methods refine cell-resolved transcriptomic inference↑ 1.6×Single-cellRank 8 Grounding biomedical LLMs in curated databases, knowledge graphs, and benchmarks↑ 1.6×ML / AI methodsRank 9 Quantitative interaction models predict community behavior, design, and dysbiosis↑ 1.6×Systems biologyRank 10 Graph and deep learning infer spatial organization from spatial transcriptomics dataTranscriptomicsRank 11 Explicit probabilistic models of nuisance structure in expression and omics measurementsProteomicsRank 12 Computational maps of cancer regulatory elements, transcripts and drug responsesML / AI methodsRank 13 Predicted and docked complexes now underpin protein-ligand affinity and pose scoringStructural bio & protein designRank 14 Computational Models Link Histone Marks and Chromatin Structure to Gene RegulationGenomicsRank 15 LLM agents for biology are validated through retrieval grounding and interactive workflowsTools & infrastructureRank 16 Ground-truth benchmarks and direct comparisons test microbiome tool reliabilityTools & infrastructureRank 17 Fine-mapping and allele-specific assays narrow candidate causal variantsGenomicsRank 18 Sequence-encoded conformational control in GPCR activation and viral protein analysisSystems biologyRank 19 Biological networks and knowledge graphs organize evidence for drug repurposingTools & infrastructureRank 20 Immune atlases map inflammatory and cytotoxic cell states across disease and agingSingle-cell
Show 18 more
Rank 21 Compact representations and sketches make large-scale genomic analysis more tractableGenomicsRank 22 Integrated pipelines and atlases unify spatial, single-cell and multi-omic profilesML / AI methodsRank 23 Coarse-grained RNA models add sugar puckering, ion and temperature effects to foldingStructural bio & protein designRank 24 Learned models are used to design and evaluate biomolecular sequencesML / AI methodsRank 25 Learned scoring of protein-protein interfaces in cryo-EM and predicted modelsStructural bio & protein designRank 26 ML rescoring and mass offset search refine peptide and PTM identification from MS/MSProteomicsRank 27 Pipelines and databases turn raw sequence data into queryable genomic resourcesTools & infrastructureRank 28 Taxon-restricted curated genome databases pair quality metrics with ecological metadataTools & infrastructureRank 29 Standardized proteomic reference resources paired with machine learning modelingProteomicsRank 30 Spike statistics to synaptic learning jointly shape neural representationSystems biologyRank 31 Computational methods resolve hard-to-call and hard-to-interpret variants in human genomesGenomicsRank 32 Multiomics decomposition and curated disease-context omics reference resourcesTools & infrastructureRank 33 Sequence-structure objectives recast as energy landscapes for protein and RNA designStructural bio & protein designRank 34 Learned image features tested against clinical criteria and explainable survival linksML / AI methodsRank 35 Curation layers are scaling LLM and ontology annotation across biological databasesTools & infrastructureRank 36 Individual-level computational estimates of brain trajectory and methylation age riskEpigenomicsRank 37 Annotation errors and stop-loss variants complicate proteogenomic interpretationGenomicsRank 38 Chemical language models and mass spectrometry expand metabolite discoveryMetabolomics