Top 40
View all ›BACKFILLED Code and industry signals are partial for backfilled days.
Rank 1
Transcriptomics
Benchmarking and new tools for inferring cell and spatial structure from transcriptomes
Proliferating spatial and single-cell tools now face systematic benchmarking alongside new methods for annotation and segmentation.
7 preprints · 13 published · ↑ 3.3× vs 4-wk avg
Rank 2 Transcriptomics
Neural models infer chromatin, transcript, and regulatory states from sequence data
Recent benchmarks and new architectures test which deep learning designs best recover chromatin contacts, accessibility, and transcript regulation.
2 preprints · 17 published · ↑ 3.1× vs 4-wk avg
Rank 3 ML / AI methods
Structural and language-model priors are being fused into protein and peptide property…
Protein language model embeddings and geometric graph encodings now constrain predictions for peptides, mutations, and membrane permeation.
2 preprints · 11 published · ↑ 2.3× vs 4-wk avg
Rank 4 Transcriptomics
Structured statistical and spatial frameworks for transcriptomic analysis
Routine transcriptomic analysis is moving beyond per-feature statistics toward network, spatial, and metabolic-modeling frameworks.
1 preprint · 8 published · ↑ 1.7× vs 4-wk avg
Rank 5 Structural bio & protein design
Learned interface representations predict TCR and antibody binding and guide design
March 2026 papers benchmark learned binding predictors and expose limits in generalization and confidence scoring.
3 preprints · 6 published · ↑ 1.7× vs 4-wk avg
Rank 6 Computational tools target noise, bias, and contamination across sequencing platforms↑ 1.6×Single-cellRank 7 Benchmarking and new tools for variant, repeat and assembly error detection↑ 1.6×GenomicsRank 8 Deep learning on incomplete omics data: interpretability, deconvolution, imputation↑ 1.6×ML / AI methodsRank 9 ML-guided generative design of peptides, RNA, and inhibitors under structural validation↑ 1.6×Structural bio & protein designRank 10 Computational methods infer 3D genome architecture from indirect or incomplete dataEpigenomicsRank 11 Graph models score drug-target and drug-disease links from networks and 3D structuresML / AI methodsRank 12 Gene and miRNA networks rank candidates by weighting evidence before proposing targetsSystems biologyRank 13 Comparative genomes link gene family expansions to adaptation and hybrid incompatibilityGenomicsRank 14 Methods borrowing cross-population and prior information for GWAS and genomic predictionGenomicsRank 15 Masked diffusion and contrastive objectives extend genomic language modelsML / AI methodsRank 16 Attention and graph models predict peptide toxicity, drug synergy, and disease genesML / AI methodsRank 17 Learned scorers adjudicate protein structure models and docking posesStructural bio & protein designRank 18 New tools assign taxa and infer relationships from proteomic and domain dataTools & infrastructureRank 19 Deep learning infers copy number and ecDNA states from sequencing and histology dataML / AI methodsRank 20 Computational inference of microbial functions, interactions, and communication networksML / AI methods
Show 12 more
Rank 21 Deep learning for drug property prediction under leakage and interpretability scrutinyStructural bio & protein designRank 22 Learned sequence models for expression, solubility, and allergen predictionML / AI methodsRank 23 Mathematical models link tumor, macrophage, and immune dynamics in breast cancerSystems biologyRank 24 Context-aware substrings and complexity metrics reframe genomic sequence analysisGenomicsRank 25 Accurate strain genotype reconstruction and phylogenetic inference from short-read dataGenomicsRank 26 Scalable algorithms for single-cell, microbiome, and metagenomic data analysisTools & infrastructureRank 27 LLM agents harmonize single-cell metadata and orchestrate reproducible analysis pipelinesTools & infrastructureRank 28 Density-conditioned deep learning recovers structural detail from experimental mapsStructural bio & protein designRank 29 Stop-codon-free regions, codon bias, and duplicated paralogs reveal gene innovationGenomicsRank 30 New methods for species trees, tree effective sample size, and mixture-model molecular…GenomicsRank 31 Multiscale and topological methods resolve structure in neural activity and morphologySystems biologyRank 32 Topology and morphometrics quantify structural change in simulated biological systemsSystems biology