Top 40
View all ›PARTIAL Some sources or summaries were unavailable for this run.
BACKFILLED Code and industry signals are partial for backfilled days.
Rank 1
ML / AI methods
Multimodal graph and language-model architectures for molecular interaction prediction
Single architectures now jointly model structural, sequential and relational representations across binding, affinity, solubility and synergy endpoints.
5 preprints · 12 published · ↑ 4.7× vs 4-wk avg
Rank 2 Tools & infrastructure
Databases link noncoding RNAs and single-cell profiles to regulatory analysis
2 preprints · 23 published · ↑ 3.5× vs 4-wk avg
Rank 3 Genomics
Curated microbial databases standardize sequence, metadata and genomes for reuse
Several 2026 NAR database releases consolidate microbiome, genome and metadata resources spanning human, plant, marine and viral niches.
20 published · ↑ 3.5× vs 4-wk avg
Rank 4 Transcriptomics
Inferring spatial domains and single-cell expression from spatial transcriptomics data
Histology-informed and graph-based models are increasingly resolving spot-level data into single-cell and three-dimensional spatial organization.
5 preprints · 16 published · ↑ 3.2× vs 4-wk avg
Rank 5 Tools & infrastructure
Curated databases link molecular glues, natural products and PTMs to drug response data
January 2026 releases extend curated drug discovery resources with structural, mechanistic and bioactivity detail for targeted therapeutics.
17 published · ↑ 3.2× vs 4-wk avg
Rank 6 Curated resources link variants to regulatory, epigenomic and haplotype-resolved context↑ 2.9×Tools & infrastructureRank 7 Curated knowledgebases expand interaction, pathway, and phenotype data across species↑ 2.9×Tools & infrastructureRank 8 Curated resources expand regulatory-element and perturbation data across species↑ 2.5×Tools & infrastructureRank 9 Graph and attention models infer regulatory networks and cancer driver genes↑ 2.4×ML / AI methodsRank 10 Curated databases expand circRNA, RNA modification, and translation annotation↑ 2.1×Tools & infrastructureRank 11 Atlases and tools link GWAS loci and QTLs to cell types, pathways, and mechanisms↑ 2.3×GenomicsRank 12 Harmonized single-cell atlases and benchmarks standardize annotation and RNA velocity↑ 2.1×Single-cellRank 13 ML and protein design models predicting release, peptide potency and sensor response↑ 1.9×ML / AI methodsRank 14 Curated metabolic, enzyme and drug databases gain programmatic, interoperable access↑ 1.8×Tools & infrastructureRank 15 Reference and population context reshape variant detection and association analysis↑ 1.9×GenomicsRank 16 Learned ensembles and trajectory analysis extend sampling of protein and RNA dynamics↑ 1.8×Structural bio & protein designRank 17 Attention and generative deep learning extract clinically framed cancer biomarkers↑ 1.8×ML / AI methodsRank 18 Hi-C-derived 3D chromosome ensembles link genome architecture to factor search and…↑ 1.8×EpigenomicsRank 19 Structure-aware graph models predict binding, GPCR activity and molecular properties↑ 1.6×ML / AI methodsRank 20 Software tools simulate and analyze sequencing, methylation, and regulatory dataTools & infrastructure
Show 20 more
Rank 21 Sequence-based models predict TCR-peptide-MHC recognition and peptide developabilityML / AI methodsRank 22 Hypergraph, deep learning, and pangenome methods close gaps in long-read assemblies↑ 1.6×GenomicsRank 23 Learned ensemble predictors mimic training-set folds and over-weight gating residues↑ 1.6×Structural bio & protein designRank 24 Sequence features and chromatin context jointly shape transcription factor bindingEpigenomicsRank 25 Explicit kinetic and thermodynamic models replace empirical tuning in biological design↑ 1.5×GenomicsRank 26 Deep models integrate multi-omics single-cell and clinical data to recover hidden…Single-cellRank 27 Small molecules tune conformational switching in neurolysin, GPCRs and PPARγStructural bio & protein designRank 28 Stochastic variability and network structure shape timing and oscillation in cellsSystems biologyRank 29 Synaptic wiring alone is proposed as a basis for assigning neuronal cell typesML / AI methodsRank 30 Cross-ancestry methods refine fine-mapping, Mendelian randomization and TWASGenomicsRank 31 Pretrained genomic and protein language models adapted for molecular prediction tasksML / AI methodsRank 32 Statistical methods recover dropout-lost expression and rare cell types in scRNA-seqSingle-cellRank 33 Simulation models couple behavior, mobility and transmission across pathogen systemsSystems biologyRank 34 Structure-informed representations are improving remote homology detection and annotationStructural bio & protein designRank 35 Nanopore direct RNA methods now resolve RNA modifications and library strandednessTranscriptomicsRank 36 Statistical frameworks for transcriptional kinetics and single-cell data benchmarkingSingle-cellRank 37 Deep learning guides ligand discovery for an RNA reader and genome-wide screeningStructural bio & protein designRank 38 Causal methods for genetic and omics data address pleiotropy and nonlinear exposure…Systems biologyRank 39 Sequence statistics and AlphaFold dynamics reveal folding constraints across proteomesGenomicsRank 40 Learned embeddings and multi-energy MS/MS advance metabolite identificationProteomics