Top 40
View all ›BACKFILLED Code and industry signals are partial for backfilled days.
Rank 1
Single-cell
Single-cell multiomics resolve cis-regulatory and eQTL effects across cell types
Large single-nucleus multiome and eQTL atlases now resolve cis-regulatory and genetic effects within brain cell types.
3 preprints · 4 published · ↑ 2.5× vs 4-wk avg
Rank 2 Structural bio & protein design
Generative and free-energy methods target peptides, antibodies, and fragment-based ligands
Sequence-only and pocket-conditioned generative models now design peptides and scFvs at scale, alongside improved fragment affinity estimation.
3 preprints · 4 published · ↑ 2.5× vs 4-wk avg
Rank 3 ML / AI methods
Graph models for molecular property prediction gain uncertainty and cross-modal inputs
Recent preprints and publications pair structural graph encoders with evidential, multiscale, or multimodal training for molecular prediction tasks.
5 preprints · 4 published · ↑ 2.2× vs 4-wk avg
Rank 4 Epigenomics
Models decode regulatory elements, replication timing and copy-number architecture
Single-cell and multiome chromatin readouts are increasingly paired with predictive and normalization models to resolve regulatory architecture.
9 preprints · 2 published · ↑ 2.2× vs 4-wk avg
Rank 5 Transcriptomics
Benchmarking and tooling for RNA sequencing methods, from small RNAs to long reads
Newer RNA-seq protocols and tools are being validated against reference or matched data, including Ultima UG100 versus NovaSeq on FFPE tissues.
3 preprints · 2 published · ↑ 2.1× vs 4-wk avg
Rank 6 Pretrained RNA and protein language models paired with structure and network features↑ 2.1×ML / AI methodsRank 7 Quantifying mosaic mtDNA, tandem repeat, and cfDNA-derived variation from sequencing reads↑ 1.9×GenomicsRank 8 Learned models span force fields, complex stoichiometry and RNA structure refinement↑ 2.0×Structural bio & protein designRank 9 Mechanistic models tie timing, migration and trait architecture to evolutionary outcomes↑ 1.8×Systems biologyRank 10 Probabilistic and dataset-aware artefact handling in metagenomic and viral sequencing↑ 1.8×GenomicsRank 11 Generative models sample biomolecular conformational ensembles from simulation and SAXS↑ 1.7×Structural bio & protein designRank 12 Computational methods extract spatial gene expression and histology features at scale↑ 1.6×TranscriptomicsRank 13 Language model embeddings and graph networks predict protein functional properties↑ 1.6×ML / AI methodsRank 14 Short-read and metagenomic tools improve plasmid recovery and classification sensitivity↑ 1.6×Tools & infrastructureRank 15 Learned transfer across bulk, spatial, and cell-line omics data for biological prediction↑ 1.5×ML / AI methodsRank 16 LLM agents drive executable omics analysis pipelines over spatial and single-cell dataML / AI methodsRank 17 Latent models infer cell fate, transitions and perturbation responsesSingle-cellRank 18 Noisy, correlated data expose identifiability limits in systems biology model fittingSystems biologyRank 19 Ancestry-aware models improve genotype inference in admixed populationsGenomicsRank 20 Integrative modeling maps dynamic assemblies in HDAC, mTORC1, PP2A, SRSF1 and Hsp70Structural bio & protein design
Show 20 more
Rank 21 Coarse-grained and enhanced-sampling models extend to nucleic acids and domain dynamicsStructural bio & protein designRank 22 LLM agents coordinate existing protein design tools into executable workflowsStructural bio & protein designRank 23 Joint image and expression latents couple histology with spatial transcriptomics across…TranscriptomicsRank 24 Metapopulation models make mobility and surveillance sensitivity inferential targetsSystems biologyRank 25 CPU and GPU acceleration of genomic variant calling, IBD detection and ASE analysisTools & infrastructureRank 26 scRNA-seq methods are being judged on robustness, transfer, and rare-cell recoverySingle-cellRank 27 Computational pipelines refine gene, ORF and exon annotation across diverse genomesGenomicsRank 28 Electrostatics as a stringent test for predicted and simulated protein structuresStructural bio & protein designRank 29 Packaged pipelines for single-cell, CLEM, spatial and force-microscopy data analysisTools & infrastructureRank 30 Learned latent representations across neural, single-cell and genomic dataSingle-cellRank 31 Pre-model data curation and validation design determine trust in omics inferenceTools & infrastructureRank 32 Language models extract enzyme and bioactivity data from literature and patentsML / AI methodsRank 33 Quantitative image analysis of mitochondria and subcellular structures across scalesML / AI methodsRank 34 Single-cell and spatial profiling resolves state tradeoffs and fate-boundary programsSingle-cellRank 35 Haplotype-level inference from sequencing, chromatin contact, and population data at scaleGenomicsRank 36 Simulated conformational ensembles map competing pathways in RNA, DNA and enzyme bindingStructural bio & protein designRank 37 Explicit probabilistic models drive genomic and biological inference tasksML / AI methodsRank 38 Pangenome graphs gain formal homology, node-count modeling, and comparison frameworksGenomicsRank 39 Computational profiling links tumor states and clonal evolution to immune shiftsGenomicsRank 40 Higher-order and dynamic connectivity metrics track brain states and disordersSystems biology