Top 40
View all ›BACKFILLED Code and industry signals are partial for backfilled days.
Rank 1
Systems biology
Network inference from omics data recovers directional and temporal molecular structure
Recent papers apply Bayesian, differential and cross-talk network models to proteomic, transcriptomic and single-cell data for target discovery.
3 preprints · 10 published · ↑ 4.3× vs 4-wk avg
Rank 2 Transcriptomics
Computational methods fuse histology, spatial and single-cell data to refine tissue maps
Spatial omics tools are proliferating, and benchmarks, simulators and metrics are now needed to test whether computed tissue measurements are faithful.
5 preprints · 11 published · ↑ 4.0× vs 4-wk avg
Rank 3 ML / AI methods
Leakage inflates enzyme-substrate scores while protein language model embeddings recur
Recent preprint and published work test whether reported ligand and peptide prediction gains survive stricter evaluation and representation choices.
1 preprint · 12 published · ↑ 3.4× vs 4-wk avg
Rank 4 Single-cell
Latent-variable models adapt to sparse, noisy single-cell and spatial measurements
Million-cell single-cell datasets and unaligned serial spatial sections now demand scalable probabilistic models that respect sparse count structure.
4 preprints · 15 published · ↑ 3.5× vs 4-wk avg
Rank 5 ML / AI methods
Sequence-based models predict protein binding, metal sites and mutation pathogenicity
Recent items pair new predictors for binding, metal sites and mutation pathogenicity with critiques of how such models are evaluated.
1 preprint · 6 published · ↑ 3.3× vs 4-wk avg
Rank 6 Deep learning models for drug-target, synergy, and repurposing prediction↑ 2.6×ML / AI methodsRank 7 Deep learning infers hidden tumor features from histology, transcriptomes and methylation…↑ 2.3×ML / AI methodsRank 8 Web and network tools extend gene-level and enrichment analysis across omics data↑ 2.1×Tools & infrastructureRank 9 Distributed cell identity encoded across single-cell and spatial transcriptomes↑ 2.0×Single-cellRank 10 Computational methods resolve copy number and clonal structure in tumor genomes↑ 1.8×GenomicsRank 11 Learned models score protein, antibody and ligand interactions from structure or sequence↑ 1.7×Structural bio & protein designRank 12 Explicit mechanistic models tie feedback and constraint to dynamics across scales↑ 2.0×Systems biologyRank 13 Computational mapping of sequence and chemistry into spaces for design and scoring↑ 1.8×Structural bio & protein designRank 14 Standardized benchmarks expose leakage and overfitting in biological ML models↑ 1.6×ML / AI methodsRank 15 Machine learning models generate and screen antimicrobial and therapeutic peptides↑ 1.5×Structural bio & protein designRank 16 Pangenome graphs resolve structural and TE variants tied to traits and adaptation↑ 1.6×GenomicsRank 17 Open-source tools target reproducibility, integration and data-specific omics analysis↑ 1.6×Tools & infrastructureRank 18 Deep models infer protein-mediated 3D chromatin contacts and loops from binding profilesEpigenomicsRank 19 Learned sequence models address data scarcity across enzymes, proteins and RNA↑ 1.5×ML / AI methodsRank 20 Learned embeddings aim to generalize biological inference across cohorts and modalities↑ 1.5×ML / AI methods
Show 20 more
Rank 21 TCR-pMHC binding models face generalization limits to unseen peptidesML / AI methodsRank 22 Deep generative models simulate and forecast single-cell transcriptional statesSingle-cellRank 23 Benchmarking conformational ensemble models of flexible multidomain proteins against SAXSStructural bio & protein designRank 24 New algorithms for nanopore read assembly, alignment, simulation and repeat analysisGenomicsRank 25 Reduced and agent-based models capture multicellular fixation, patterning, and fate↑ 1.6×Systems biologyRank 26 Alignment-free k-mer features resolve variant and class differences in genomic sequencesGenomicsRank 27 Synaptic tuning, energy cost, and connection type bound neural information↑ 1.5×Systems biologyRank 28 New tools map RNA-binding protein sites and benchmark RNA quantification accuracyGenomicsRank 29 Model-based GWAS and ancestry methods extend to imaging and multi-phenotype outcomesGenomicsRank 30 Reaction network and surrogate models predict and steer gene expression behaviorSystems biologyRank 31 Machine learning and Bayesian methods sharpen genetic association and predictionGenomicsRank 32 Learned models stress-test binding predictions and steer molecular generationStructural bio & protein designRank 33 Pipelines and models couple cell painting morphology to transcriptomic and spatial dataTools & infrastructureRank 34 Tools assigning sequencing reads to taxa, transcripts and repeats are benchmarkedTranscriptomicsRank 35 Metabolic network models link microbial interactions to host disease and therapySystems biologyRank 36 Scored comparisons reveal how complexes, secretomes and accessory genomes diversifyGenomicsRank 37 Epigenomic and 3D genome data are being modeled to infer regulatory architectureEpigenomicsRank 38 Experimental data arbitrate computational RNA structure and potential predictionsStructural bio & protein designRank 39 Learned scoring and active learning accelerate docking-based virtual screeningStructural bio & protein designRank 40 Repository-scale and benchmarked pipelines widen computational peptide identificationProteomics