Top 40
View all ›BACKFILLED Code and industry signals are partial for backfilled days.
Rank 1
Single-cell
Explicit latent structure in omics models handles missing, unpaired, and spatial data
Recent preprints and papers embed latent variable structure, factorisations, and conditional generative components directly into omics models to handle…
11 preprints · 2 published · ↑ 2.5× vs 4-wk avg
Rank 2 Epigenomics
Models map chromatin context and 3D contacts onto expression and tissue architecture
Recent preprints pair 1D epigenetic tracks and 3D chromatin interactions with deep learning to predict expression and map spatial domains.
7 preprints · 3 published · ↑ 2.4× vs 4-wk avg
Rank 3 Metabolomics
Computational inference from tandem mass spectra for metabolite structure and identity
Learned and search-based models now pair spectral comparison with de novo structure proposal for metabolite annotation across public MS/MS data.
4 preprints · 2 published · ↑ 2.4× vs 4-wk avg
Rank 4 Tools & infrastructure
Computational pipelines for bacterial genomics, viral profiling and amplicon analysis
Recent work targets error-prone amplicon diversity estimates and reproducible pathogen genomics workflows across sequencing platforms.
5 preprints · 3 published · ↑ 2.3× vs 4-wk avg
Rank 5 Structural bio & protein design
Coarse pocket and deep-learning scoring models predict protein-ligand binding affinity
Recent preprints replace full atomic inputs with coarse pocket or large DNA-encoded library-trained representations for faster protein-ligand affinity…
4 preprints · 1 published · ↑ 2.0× vs 4-wk avg
Rank 6 New algorithms speed phylogenetic placement and phase viral haplotypes from long reads↑ 1.8×GenomicsRank 7 Interpretable deep models probe protein, RNA, and biomedical graph relations↑ 1.6×ML / AI methodsRank 8 Sequence language models are evaluated on structure, fitness and interaction outcomes↑ 1.6×ML / AI methodsRank 9 Learned models predict biomolecular structure, peptide properties and retention times↑ 1.6×ML / AI methodsRank 10 Harmonized metadata resources make heterogeneous public omics records comparable and…↑ 1.8×Tools & infrastructureRank 11 Computational tools for proteomics quantification and structure-based motif matching↑ 1.6×ProteomicsRank 12 LLM agents add pharmacological reasoning and multi-step workflows to molecular designML / AI methodsRank 13 Learned and compressed feature spaces are used to classify and curate cryo-EM particlesStructural bio & protein designRank 14 Training dynamics and context length shape LLM alignment with brain activity↑ 1.6×ML / AI methodsRank 15 Computational frameworks resolve senescent and hepatic cell states from single-cell dataSingle-cellRank 16 Machine learning models predict RNA regulatory sites from sequence and expression dataML / AI methodsRank 17 Generative evolution models tested on protein fitness and alignment diversityStructural bio & protein designRank 18 Natural-language interfaces make omics and biochemical databases interactively queryableTools & infrastructureRank 19 Generative models steer biomolecular design via alignment, energy biasing, and samplingStructural bio & protein designRank 20 Structured models infer expression, lineage, and aging states from omics dataSingle-cell
Show 17 more
Rank 21 Computational mapping of regulatory variants across epigenomic and population scalesEpigenomicsRank 22 Biological and acquisition structure constraining learned imaging representationsML / AI methodsRank 23 Packaged tools for epidemic simulation, ancestral graphs and trajectory analysisTools & infrastructureRank 24 Statistical methods for RNA-seq power, meta-analysis, batch effects and DE testingTools & infrastructureRank 25 Validation pitfalls inflate or obscure predictive performance in computational biologyML / AI methodsRank 26 Machine learning infers biological age, circadian time and post-mortem intervalEpigenomicsRank 27 LLM agents and sequence-conditioned models grounded in single-cell biological evidenceSingle-cellRank 28 Learned models predict promoter activity and forecast single-cell state transitionsSingle-cellRank 29 Machine learning on single-cell atlases nominates IBD and SLE disease targetsSingle-cellRank 30 Computational structure-based methods target kinase selectivity and P-gp inhibitionStructural bio & protein designRank 31 New tools infer population size, merge DNA motifs, and model microbiome disease featuresTools & infrastructureRank 32 Cloud and browser platforms packaging AI-driven molecular design and simulationTools & infrastructureRank 33 Network and co-essentiality evidence guides computational gene prioritizationTools & infrastructureRank 34 Ion mobility mass cytometry resolves metabolites in single aging mouse liver cellsMetabolomicsRank 35 Sequence models and genome mining map Cas9 PAMs, plasmids and MRSA epistasisGenomicsRank 36 Noise-robust MRI connectivity reveals hidden inter-subject variabilitySystems biologyRank 37 Tissue-wide planar cell polarity emerges from local signals in cellular Potts modelsSystems biology