Top 40
View all ›BACKFILLED Code and industry signals are partial for backfilled days.
Rank 1
Single-cell
Computational methods resolve cell-type-specific programs in single-cell and spatial data
Sparse, heterogeneous single-cell and spatial datasets are prompting methods that model cell-type specificity and uncertainty directly.
11 preprints · 6 published · ↑ 4.4× vs 4-wk avg
Rank 2 ML / AI methods
Deep learning predicts enzyme reactions, binding sites, and drug-target affinity
Protein language models, graph networks, and meta-learning are being applied to enzyme, peptide, and drug-target prediction tasks.
6 preprints · 7 published · ↑ 3.4× vs 4-wk avg
Rank 3 Genomics
Neural and probabilistic models resolve population-scale genetic variation
Long-read, pangenome, and biobank datasets now support new methods for phasing, structural variant genotyping, ancestry inference, and fine-mapping.
4 preprints · 2 published · ↑ 3.0× vs 4-wk avg
Rank 4 Structural bio & protein design
Structure-based binding prediction guides epitope, miniprotein, and cytokine design
Predicted affinity shifts from phosphorylation and AlphaFold3-guided design are now moving toward experimental validation via ELISA and PBMC assays.
3 preprints · 3 published · ↑ 3.0× vs 4-wk avg
Rank 5 ML / AI methods
Language and sequence embeddings fused with graph models for drug response prediction
Recent preprints and publications pair pretrained text, sequence and molecular embeddings with graph or Mamba encoders for drug-related prediction tasks.
1 preprint · 5 published · ↑ 3.0× vs 4-wk avg
Rank 6 Computational curation resolves collapsed homeologs in chromosome-scale assemblies↑ 2.7×GenomicsRank 7 Image-derived metrics quantify motion, organization and aggregation in cells↑ 2.7×Tools & infrastructureRank 8 Deep learning infers cell morphology, lineage and state from imaging and single-cell data↑ 2.7×Single-cellRank 9 Workflows standardize bioimage and single-cell analysis from raw inputs to results↑ 2.7×Tools & infrastructureRank 10 Computational models infer regulatory state from epigenomic and sequence data↑ 2.5×EpigenomicsRank 11 Context-resolved regulatory variant mapping across tissues, cell types, and aging↑ 2.5×EpigenomicsRank 12 Pretrained models are reused across omics and imaging modalities under limited labels↑ 2.4×ML / AI methodsRank 13 Model assumptions shape what evolutionary and kinetic analyses recover↑ 2.4×Systems biologyRank 14 Faster and benchmarked analysis methods for calcium and single-molecule microscopy↑ 2.3×ML / AI methodsRank 15 Model-based classifiers and pipelines for microbial strain and species profiling↑ 2.1×Tools & infrastructureRank 16 Pretrained sequence models are being diagnosed for what their representations encode↑ 2.1×ML / AI methodsRank 17 Defective genomes and sequencing artefacts are being masked in viral consensus analyses↑ 2.0×GenomicsRank 18 Coarse-grained and generative models sample biomolecular ensembles and condensation↑ 2.0×Structural bio & protein designRank 19 MD maps conformational ensembles in ribozyme catalysis, actin Pi release, kinesin↑ 2.0×Structural bio & protein designRank 20 Attractor and wave dynamics in cellular and network models of biology↑ 2.0×Systems biology
Show 20 more
Rank 21 Reproducible DIA proteomics workflows and scRNA-seq differential expression benchmarks↑ 2.0×Single-cellRank 22 Structured-population models reveal how sampling and assumptions shape pathogen inference↑ 1.9×Systems biologyRank 23 GPU-scale acceleration of biomolecular structure prediction and cryo-EM projection↑ 2.0×Structural bio & protein designRank 24 Pretrained sequence language models adapted for protein and genomic prediction↑ 1.8×ML / AI methodsRank 25 Computational calling of genomic events from long-read, ctDNA and single-cell sequencing…↑ 1.9×Single-cellRank 26 When structure-based recovery of cellular dynamics models remains reliable↑ 1.8×Single-cellRank 27 Deep models transfer information across contexts to predict unmeasured perturbations↑ 1.6×ML / AI methodsRank 28 Modeling detectability, missingness and dose-response in omics measurement↑ 1.6×ML / AI methodsRank 29 Joint optimization of synthesizability and binding affinity in fragment-based design↑ 1.5×Structural bio & protein designRank 30 Transcriptome-based inference of RNA regulators, from splicing factors to RBP bindingTranscriptomicsRank 31 Models show lysis timing, migration, and coevolution reshape pathogen and phage dynamics↑ 1.5×Systems biologyRank 32 Generative and data-constrained models sample biomolecular conformational ensemblesStructural bio & protein designRank 33 Graph-based and outlier-trimmed methods for spatial and single-cell omics dataGenomicsRank 34 Scalable software for accelerating phylogenetic likelihoods, variant calling, and…Tools & infrastructureRank 35 Computational prediction of GPCR-G protein and TRP channel interaction partnersStructural bio & protein designRank 36 Latent models infer neural dynamics from RNNs, populations and whole-brain data↑ 1.5×Systems biologyRank 37 Machine learning resolves antigen-specific and clonal lymphocyte receptor structureSingle-cellRank 38 Foundation segmentation models automate morphological measurement from microscopy imagesTools & infrastructureRank 39 Learned protein representations extend function, site, and stoichiometry predictionStructural bio & protein designRank 40 Quantitative models infer tumor subpopulation competition and drug resistance mechanismsSystems biology