Top 40
View all ›BACKFILLED Code and industry signals are partial for backfilled days.
Rank 1
ML / AI methods
Machine learning links molecular and clinical features to predicted cancer drug response
scDrugMap benchmarks foundation models across 495,000 cells while GDSC-based fusion models predict kinase inhibitor sensitivity in cell lines.
4 preprints · 4 published · ↑ 2.9× vs 4-wk avg
Rank 2 ML / AI methods
Biological priors shape embeddings for single-cell, spatial, and clinical omics
Foundation models and generative methods now extend constrained representations from scRNA-seq to histology, spatial omics, and longitudinal transcriptomics.
13 preprints · 1 published · ↑ 2.2× vs 4-wk avg
Rank 3 ML / AI methods
ML drug discovery models add explicit uncertainty and error control to predictions
Screening and ADMET decisions increasingly depend on calibrated confidence rather than point predictions alone.
3 preprints · 4 published · ↑ 2.0× vs 4-wk avg
Rank 4 Structural bio & protein design
Residue-resolved energetics and side-chain flexibility in ligand recognition
Recent preprints and a 2025 GPCR study apply flexible docking, perturbation scanning, and graph learning to assign binding determinants at residue level.
2 preprints · 1 published · ↑ 2.0× vs 4-wk avg
Rank 5 Systems biology
Inference frameworks link gut microbes and metabolites to host and disease outcomes
Recent preprints package double machine learning, penalized regression and graph neural networks into R tools for microbiome multi-omics.
6 preprints · 3 published · ↑ 1.8× vs 4-wk avg
Rank 6 Assemblies link genome architecture to size, rearrangement and structural variation↑ 1.9×GenomicsRank 7 Computational methods trace atoms, stereochemistry and assembly complexity in metabolites↑ 1.9×MetabolomicsRank 8 Software tools for benchmarking, standardising and visualising omics analysis outputs↑ 1.8×Tools & infrastructureRank 9 Coarse-grained simulations define state ensembles for chromatin and DEAD-box proteins↑ 1.8×Structural bio & protein designRank 10 Task-specific k-mer, minimizer, and graph indexing for pangenome and ancient DNA↑ 1.6×GenomicsRank 11 Single-cell methods link genetic regulation and cell communication to disease biology↑ 1.6×Single-cellRank 12 Curated data and generative models steer conditional design of proteins and molecules↑ 1.6×Structural bio & protein designRank 13 Learned models for cell segmentation, classification, and super-resolution imaging↑ 1.5×ML / AI methodsRank 14 Learned latent representations decode fMRI, iEEG, LFP and multiscale neural activity↑ 1.5×ML / AI methodsRank 15 Haplotype and recombination structure encoded in genomic analysis software↑ 1.6×GenomicsRank 16 Transcriptomic and multi-omic models predicting therapeutic response across tissuesSingle-cellRank 17 LLM agents direct bioinformatics tools for scRNA-seq annotation and NGS interpretationSingle-cellRank 18 Augmenting scarce labeled peptide data with generative models for classificationML / AI methodsRank 19 Benchmarked method choices in GPCR activation and drug residence time predictionStructural bio & protein designRank 20 Pretrained sequence embeddings with ensemble heads predict genomic and protein sitesML / AI methods
Show 20 more
Rank 21 Genome-scale models chart bacterial metabolism and gene function across strainsSystems biologyRank 22 Computational frameworks link connectivity, plasticity, and neural responsesML / AI methodsRank 23 Benchmarks assess generalizability of perturbation response models across datasetsSystems biologyRank 24 GWAS loci are translated into genes, cell types, and shared genomic factorsGenomicsRank 25 Evidence beyond raw calls refines variant classification and mutational signaturesGenomicsRank 26 RBP binding links exon inclusion to histone marks and NMD efficiency in tumorsEpigenomicsRank 27 Known-truth simulators and benchmarks expose limits of sequencing analysis toolsTools & infrastructureRank 28 In silico screen and cross-species genomics link regulators to chromatin and G4 potentialGenomicsRank 29 Benchmarked feature scoring extends from bulk expression to isoform and chromatin dataSingle-cellRank 30 Genomic architecture and chromatin dynamics underlying plant abiotic stress responsesGenomicsRank 31 Predictor reliability varies for disordered multimers and amyloid polymorphsStructural bio & protein designRank 32 Learned estimators recover structure from noisy single-cell and spatial expression dataTranscriptomicsRank 33 Methods sharpen cell-type resolution in single-cell and spatial transcriptomic analysisSingle-cellRank 34 Learned samplers are validated against physical energies and free-energy landscapesStructural bio & protein designRank 35 Computational inference of protein localization and secretory pathway regulationProteomicsRank 36 Methods partition genetic, regulatory and interaction effects on complex traitsSystems biologyRank 37 Learned and sparse models infer nonlinear fitness landscapes and population dynamicsGenomicsRank 38 Structure-conditioned generative design extends to nucleases and antibody repertoiresStructural bio & protein designRank 39 Multi-layer computational prioritization of tumor neoantigens and pHLA off-target riskTools & infrastructureRank 40 Neural networks constrained by mechanistic structure fit pathway and trajectory dynamicsSystems biology