Top 40
View all ›BACKFILLED Code and industry signals are partial for backfilled days.
Rank 1
ML / AI methods
Benchmarks and uncertainty-aware models for spatial transcriptomics and multi-omics
Benchmarks now expose method variability in spatial omics, pushing toward calibrated uncertainty and multimodal integration across domains and genes.
8 preprints · 12 published · ↑ 3.3× vs 4-wk avg
Rank 2 Transcriptomics
Deep neural architectures address single-cell and spatial transcriptomics bottlenecks
Hybrid architectures pairing graph diffusion, attention, or language model embeddings are now benchmarked against specific single-cell bottlenecks.
2 preprints · 16 published · ↑ 3.0× vs 4-wk avg
Rank 3 ML / AI methods
Protein language and structure models augmented with chemical and evolutionary priors
Recent preprints and publications test whether added priors and confidence scores reflect true binding specificity in predicted protein complexes.
9 preprints · 8 published · ↑ 2.9× vs 4-wk avg
Rank 4 Systems biology
Network-aware models extend transcriptomic inference beyond gene-level testing
Spatial and single-cell data are driving methods that model network context and spatial structure beyond standard differential expression.
5 preprints · 8 published · ↑ 2.3× vs 4-wk avg
Rank 5 Genomics
Error-focused benchmarking of structural, copy number and tandem repeat variant callers
Long-read and large-cohort sequencing now support systematic assessment of variant callers against defined error modes.
3 preprints · 8 published · ↑ 2.0× vs 4-wk avg
Rank 6 Learned interface models now drive binding prediction and generative protein design↑ 1.9×Structural bio & protein designRank 7 Graph neural models for drug repurposing, synergy, and protein-ligand modelling↑ 1.9×ML / AI methodsRank 8 Methods reconstruct cellular and spatial structure from spatial transcriptomics↑ 1.9×TranscriptomicsRank 9 Learned representations drive molecular interaction and activity prediction models↑ 1.7×ML / AI methodsRank 10 Deep models predict perturbation, variant, and docking effects with transfer guarantees↑ 1.7×ML / AI methodsRank 11 Association and prediction methods extend across populations, priors and omics layers↑ 1.7×GenomicsRank 12 Protocol and metric choices shape conclusions drawn from single-cell analyses↑ 1.7×Single-cellRank 13 Learned and experimental models map noncoding variants onto 3D regulatory contacts↑ 1.6×EpigenomicsRank 14 Deep learning evaluated under constrained biomedical data: patches, few-shot, anomaliesML / AI methodsRank 15 Machine learning infers copy number and tumor content from indirect cancer measurementsML / AI methodsRank 16 Generative and algorithmic models design RNA, mRNA, enzyme domains, and peptidesStructural bio & protein designRank 17 Deep models predict and generate peptides and molecules; benchmark leakage exposedML / AI methodsRank 18 Comparative genomes link gene family expansions to adaptation and hybrid incompatibilityGenomicsRank 19 Learned models predict protein and RNA structural and functional featuresStructural bio & protein designRank 20 New metagenomic tools resolve strain genotypes, assemblies, pseudogenes and phylogeniesGenomics
Show 20 more
Rank 21 Single mechanisms decide persistence, extinction, or saturation in biological dynamicsSystems biologyRank 22 Learned latent models infer hidden cell-state dynamics from single-cell count dataSingle-cellRank 23 LLM agents constrained to versioned tools for single-cell and spatial analysisTools & infrastructureRank 24 Mathematical and causal models dissect tumor-immune dynamics in breast cancer therapySystems biologyRank 25 Software tools operationalize genotype-to-phenotype prediction and variant interpretationTools & infrastructureRank 26 Computational inference of microbial function and growth from community dataSystems biologyRank 27 Simulations resolve conformational ensembles of flexible, aggregation-prone biomoleculesStructural bio & protein designRank 28 GPU and algorithmic redesign accelerate single-cell, microbiome, and RNA-seq analysisTools & infrastructureRank 29 New statistical methods for species trees, Bayesian sampling, and molecular datingGenomicsRank 30 Pairing mechanistic and learned models with single-cell data to infer cell dynamicsSingle-cellRank 31 Spectra and peptide confidence drive microbial identity inference in proteomicsProteomicsRank 32 Coarse-grained simulations link sequence rules to multiphase condensate morphologySystems biologyRank 33 Sequence composition and length structure vary across genomic scalesGenomicsRank 34 Sequence constraints guide proto-gene origin and paralog divergenceGenomicsRank 35 Python and Spark tools extract genotype and phylogenetic structure from sequencing dataTools & infrastructureRank 36 Deep learning and flexible alignment accelerate receptor-aware virtual screeningStructural bio & protein designRank 37 Statistical and spectral methods for quantifying proteins and single-cell fluorescence…ProteomicsRank 38 Quantum-accurate energy functions and charge models for biomolecular simulationStructural bio & protein designRank 39 Separating stimulus- and behavior-linked components from neural activity dynamicsSystems biologyRank 40 Learned latent spaces condition protein structure, conformation and sequence generationStructural bio & protein design