Top 40
View all ›BACKFILLED Code and industry signals are partial for backfilled days.
Rank 1
ML / AI methods
Graph neural networks predict drug, protein and host-pathogen interactions
Graph-based deep learning is being combined with multimodal fusion and explainability to predict biological and chemical interactions for drug and target…
6 preprints · 6 published · ↑ 4.0× vs 4-wk avg
Rank 2 ML / AI methods
Soil plasmidomes and Hi-C virus-host benchmarks extend microbial mobile-gene mining
Large plasmid and virus-host resources now come paired with explicit benchmarks or ecological context, tying mobile genes to their hosts.
3 preprints · 4 published · ↑ 3.3× vs 4-wk avg
Rank 3 Single-cell
Deep models align histology, spatial transcriptomics, and unpaired single-cell data
Preprints this month pair histology, spatial transcriptomics, proteomics, and unpaired single-cell data with deep alignment and translation models.
6 preprints · 3 published · ↑ 3.0× vs 4-wk avg
Rank 4 Structural bio & protein design
Sequence-based deep learning predicts protein binding and functional properties
Learned protein language model embeddings now support sequence-only prediction of binding affinity, interactions, and allergenicity.
10 preprints · 4 published · ↑ 3.1× vs 4-wk avg
Rank 5 Single-cell
Temporally resolved single-cell data fitted to dynamical models infer cell behavior
Minute-scale scRNA-seq and multi-omic velocity models now resolve dynamic cell-state transitions beyond static embeddings.
9 preprints · 1 published · ↑ 2.7× vs 4-wk avg
Rank 6 Learned sequence and protein-language scores now prioritize non-coding and rare variants↑ 2.7×GenomicsRank 7 Sampling frames and primer design decide which microbial lineages enter analysis↑ 2.7×GenomicsRank 8 Interaction-aware neural models decode EEG, SEEG, ECoG and connectome structure↑ 2.1×ML / AI methodsRank 9 Computational methods infer cell types and spatial expression from transcriptomic data↑ 2.0×TranscriptomicsRank 10 Multi-omic models infer disease-specific metabolic states and regulatory circuits↑ 2.0×Systems biologyRank 11 Foundation models predict regulatory activity and 3D chromatin from DNA sequence↑ 2.0×ML / AI methodsRank 12 Computational tools map non-canonical RNA and DNA structures and gene loss at scale↑ 2.0×Tools & infrastructureRank 13 Learned models and benchmarks for genomic variation, expression and mutation rates↑ 1.7×GenomicsRank 14 Population-scale genomes resolve diversity, domestication and inherited methylation↑ 1.8×GenomicsRank 15 Gapless and population-scale references set new baselines for genomic analysis↑ 1.8×GenomicsRank 16 Generative models now design ligands, peptides and binders conditioned on 3D structure↑ 1.5×Structural bio & protein designRank 17 Sequence and nucleosomes govern chromatin structure and methylation-dependent binding↑ 1.8×EpigenomicsRank 18 Frameworks couple mechanistic models with ML calibration and noise-aware inference↑ 1.6×ML / AI methodsRank 19 Causal and enrichment criteria are qualifying network-derived biological predictions↑ 1.5×Systems biologyRank 20 Energetic and dynamic characterization of protein structures and ligand bindingStructural bio & protein design
Show 20 more
Rank 21 Language models guide generative redesign of proteins and de novo genesStructural bio & protein designRank 22 Generative and transport models infer single-cell dynamics and perturbation responsesSingle-cellRank 23 Membrane geometry and gradients set local-rule-driven oscillation and ion homeostasisSystems biologyRank 24 Machine learning and inference applied directly to raw single-trace and image dataML / AI methodsRank 25 Genomic language models learn adaptive tokens and scale context inferenceML / AI methodsRank 26 LLM and network methods link longitudinal blood, gene literature, and therapeutic dataSystems biologyRank 27 Near-T2T cattle genome, Y-haplogroup calling workflow, and standardized assembly metricsGenomicsRank 28 SIMD, GPU and diagonal-partitioning methods speed up sequence alignment and motif analysisTools & infrastructureRank 29 Deep models predict HIV diagnoses, protein succinylation, and RNA methylation sitesML / AI methodsRank 30 Dynamics analysis and NMR-guided sampling sharpen protein conformational ensemblesStructural bio & protein designRank 31 Open-source R workflows span information-theoretic, spatial, and multi-omics analysesTools & infrastructureRank 32 Explicit computational decision rules encoded in RNA, DNA, and splicing analysis toolsTranscriptomicsRank 33 Symbolic and thermodynamic models link network structure to reaction dynamicsSystems biologyRank 34 Omitted epistasis biases GWAS nulls; multivariate and metabolite QTL maps expand lociGenomicsRank 35 Network models separate direct from indirect gene transfers and refine amplicon unitsGenomicsRank 36 Simulations and assays map lipid transfer and pore routes in membrane remodelingStructural bio & protein designRank 37 Learning rules and circuit architecture are being fit to behavior and neural dynamicsSystems biologyRank 38 Open-source LLM agents map natural-language queries onto curated biological databasesTools & infrastructureRank 39 Delay-coupled oscillators, network size, and frequency allocation shape synchrony regimesSystems biologyRank 40 Noise, dispersal, and nutrient limits reshape predator-prey population dynamicsSystems biology