Top 40
View all ›BACKFILLED Code and industry signals are partial for backfilled days.
Rank 1
ML / AI methods
Joint learning across coupled modalities and tasks in neurological and CNS applications
Recent work couples heterogeneous data sources and prediction targets within single models for neurological and CNS tasks.
3 preprints · 6 published · ↑ 3.2× vs 4-wk avg
Rank 2 Single-cell
Learned models reconstruct single-cell state dynamics and spatial niches at atlas scale
Atlas-scale corpora of 28.5M to about 100M cells now support learned models of cell state, dynamics, and perturbation response.
10 preprints · 5 published · ↑ 2.8× vs 4-wk avg
Rank 3 Genomics
Multilocus models aggregate variant, methylation, and single-cell chromatin data
Single-cell chromatin annotations, multimodal cfDNA features and population-informed calling are converging on richer genomic architecture models.
3 preprints · 6 published · ↑ 2.7× vs 4-wk avg
Rank 4 Single-cell
Computational methods infer spatial and single-cell structure from transcriptomic data
Recent benchmarks and frameworks test whether inference methods recover 3D, perturbation, and composition structure beyond what raw counts or sections report.
3 preprints · 10 published · ↑ 2.4× vs 4-wk avg
Rank 5 Tools & infrastructure
Reproducible tools for reference-aware genome assembly and haplogroup inference
Containerized assembly pipelines and probabilistic haplogroup scoring now extend reference-aware methods to reproducible, open software.
2 preprints · 9 published · ↑ 2.2× vs 4-wk avg
Rank 6 Attention over pretrained embeddings extends biomolecular property and structure…↑ 1.9×ML / AI methodsRank 7 Comparative genomics resolves lineage history via ancestral linkage and duplication↑ 1.9×GenomicsRank 8 Learned scoring advances RNA structure probabilities and protein sequence alignment↑ 1.6×ML / AI methodsRank 9 New statistical methods for methylation, distributed reference charts, and GWAS censoringEpigenomicsRank 10 Generative models and property predictors guide de novo peptide and protein designStructural bio & protein designRank 11 Models couple sequence learning and simulation to resolve multiscale protein biophysics↑ 1.5×Structural bio & protein designRank 12 Deep learning models decode proteins, peptides and transcription factors from biological…ML / AI methodsRank 13 Open-source packages wrap multi-omics, microbiome and single-cell analysesTools & infrastructureRank 14 Open-source tools for simulating regulatory circuits, sequences, and molecular systemsTools & infrastructureRank 15 Priors and statistical weighting guide omics classification and feature selectionML / AI methodsRank 16 Deep learning across RNA structure, design, and nanopore current interpretationML / AI methodsRank 17 Targeted protein prediction tools wrap learned components into workflow-level pipelinesTools & infrastructureRank 18 Analysis tools resolving mixed and repetitive sequences in long-read genomic dataGenomicsRank 19 Computational tools predict aptamer, peptide, RNA-ligand and siRNA activityTools & infrastructureRank 20 Graph learning predicts drug synergy, ncRNA-drug links and synthetic lethalityML / AI methods
Show 17 more
Rank 21 Benchmarking and fine-tuning LLMs for biomedical relation and entity tasksML / AI methodsRank 22 Computable models of cell migration, polarity, actin assembly, and EGFR signalingSystems biologyRank 23 Epigenetic feedback and miRNA competition reshape stochastic gene regulatory dynamicsSystems biologyRank 24 New omics computing tools span survival modeling, cell annotation and isoform analysisSingle-cellRank 25 Free energy and metadynamics methods probe PARP selectivity and MAX conformationsStructural bio & protein designRank 26 Intrinsic DNA sequence motifs predict replication origins and nucleosome occupancyEpigenomicsRank 27 Genome-resolved analyses couple uncultured prokaryotes and macroalgae to environmental…GenomicsRank 28 Benchmarks test co-folding models AF3, Boltz-2 and Chai-1 against docking for ligandsStructural bio & protein designRank 29 Open tools now resolve Plasmodium genomic variation and vivax recurrence statesTools & infrastructureRank 30 Latent dynamical models extend to clinical iEEG, spike generation, and latent validationSystems biologyRank 31 Learned priors and acquisition control advance cryo-EM and electron microscopyStructural bio & protein designRank 32 Coupled behavioral, emulation, and agent-based models of pathogen transmissionSystems biologyRank 33 LLMs generating bioinformatics plots and algorithms with verifiable, checkable outputsTools & infrastructureRank 34 Evolutionary simulators gain multi-chromosome, indel, and migration-aware modelingTools & infrastructureRank 35 Molecular networks learning and memory; generative design of transcription activator…Systems biologyRank 36 Learned and kinetic models infer hidden temporal structure from transcriptomic readoutsSystems biologyRank 37 Cell-type-resolved mouse atlas links expression variation across founder strainsGenomics