Top 40
View all ›BACKFILLED Code and industry signals are partial for backfilled days.
Rank 1
Single-cell
Computational methods are standardizing single-cell and spatial transcriptomics analysis
Recent benchmarks of 29 RNA velocity algorithms and of preprocessing and integration methods test single-cell tools on standardized datasets.
3 preprints · 11 published · ↑ 4.3× vs 4-wk avg
Rank 2 Epigenomics
Multimodal models fuse chromatin, methylation and variant features for genomic inference
Functional annotations from single-cell, single-molecule and population data are now being combined directly in variant and methylation models.
5 preprints · 11 published · ↑ 2.9× vs 4-wk avg
Rank 3 Tools & infrastructure
Containerized pipelines automate chromosome-scale assembly from HiFi and Hi-C reads
Reference-aware tools for assembly, contamination removal, and bacterial genome clustering are now published as containerized, reproducible software.
1 preprint · 9 published · ↑ 2.7× vs 4-wk avg
Rank 4 Single-cell
Learned models of cell state for dynamics, regulatory networks and atlas integration
Atlas-scale data, including 28.5M healthy cells and a 100-million-cell integration target, motivate learned models of cellular state.
6 preprints · 4 published · ↑ 1.9× vs 4-wk avg
Rank 5 ML / AI methods
Adapting and benchmarking language models against curated biomedical ground truth
Open fine-tuning pipelines and curated annotated corpora now allow LLM performance to be tested against explicit biomedical references.
1 preprint · 6 published · ↑ 1.7× vs 4-wk avg
Rank 6 Benchmarks expose setting-dependent limits of learned protein-ligand binding and docking↑ 1.6×Structural bio & protein designRank 7 Methods decode spatial domains, isoforms, and 3D structure from spatial transcriptomics↑ 1.6×TranscriptomicsRank 8 Reusable pipelines and web platforms now package omics analysis steps for distribution↑ 1.6×Tools & infrastructureRank 9 LLM agents anchor biological outputs to checkable references and executable validation↑ 1.6×ML / AI methodsRank 10 Coupled multimodal models address CNS disease detection and prediction tasks↑ 1.6×EpigenomicsRank 11 Graph pangenomes and gap-free assemblies resolve lineage-level structural variation↑ 1.5×GenomicsRank 12 Generative and interpretable models link histology, imaging and omics for biomarkersML / AI methodsRank 13 Computational design and prediction tools for RNA-directed therapeuticsML / AI methodsRank 14 Diverse cohorts and inversions refine how genetic variants map to complex traits↑ 1.6×GenomicsRank 15 Attention-based encoders predict protein, nucleic acid, and phage-host interactionsML / AI methodsRank 16 Graph-based representation learning for drug, protein, and molecular property predictionML / AI methodsRank 17 Deep learning models predict essential genes, drug synergy, enhancers, and PTM sitesML / AI methodsRank 18 Biological priors and network structure guide omics driver and interaction inferenceSystems biologyRank 19 Slow regulatory variables and noise reshape stochastic cell-state dynamicsSystems biologyRank 20 Protein language model embeddings predict antibody, epitope and peptide-MHC propertiesML / AI methods
Show 18 more
Rank 21 Missingness-aware metrics and Hardy-Weinberg testing refine population genomic inferenceGenomicsRank 22 Computational methods resolve protein dynamics and electronic structure across timescalesStructural bio & protein designRank 23 DNA sequence encodes nucleosome classes that shape transcription factor bindingEpigenomicsRank 24 AlphaFold ensembles mimic training structures and miss alternative conformational statesStructural bio & protein designRank 25 Single-cell models resolve isoform, splicing, state and tissue-specificity variationSingle-cellRank 26 New tools link metabolic network visualization to metabolomic and chemical genomic dataSystems biologyRank 27 Modular Software for Structure Prediction, Network Inference, and Sequence SimulationTools & infrastructureRank 28 Computable models link polarity, protrusion, and invasion in migrating cellsSystems biologyRank 29 Simulation-based inference from sequences is being made faster and more scalableGenomicsRank 30 New valence geometry validation and 3D analysis tools for nucleic acid structuresStructural bio & protein designRank 31 Inferring time-linked variables from nascent, untimed, or mislabeled transcriptomic dataSystems biologyRank 32 Proteome-wide annotation and structure-based tools advance protein function inferenceProteomicsRank 33 Coalescent and Bayesian methods extend to migration, gene flow, and admixtureGenomicsRank 34 Portable web tools now visualize read alignments, pedigrees and outbreak genomic dataTools & infrastructureRank 35 Long-read assembly errors and k-mer diagnostics expose limits in repeats and pangenomesGenomicsRank 36 Evolutionary and structural constraints define conserved functional sites in proteinsStructural bio & protein designRank 37 Provenance-tracked analysis coupled to curated biological data resourcesTools & infrastructureRank 38 Targeted pipelines for bacterial, pathogen and phage protein analysisTools & infrastructure