Top 40
View all ›BACKFILLED Code and industry signals are partial for backfilled days.
Rank 1
Transcriptomics
Computational inference of tumor features from histology, long reads, and cfDNA
Machine learning on histology, read-level methylation, and cell-free data now recovers tumor features from indirect measurements.
10 preprints · 2 published · ↑ 2.1× vs 4-wk avg
Rank 2 ML / AI methods
Benchmarking deep and foundation models when labeled data or input scale is limited
Foundation models and self-supervised encoders are now benchmarked head-to-head against task-specific networks in small-data biomedical imaging and omics.
10 preprints · ↑ 1.9× vs 4-wk avg
Rank 3 Systems biology
Mechanistic models of tumor-immune interactions and treatment scheduling in cancer
Mechanistic models now resolve macrophage phenotype reprogramming, NK cytotoxicity, and chemotherapy scheduling against tumor-immune data.
6 preprints · 1 published
Rank 4 ML / AI methods
ML models predict protein palmitoylation, LLPS drivers and disease transcriptome states
Protein language model embeddings and multi-branch deep networks now drive predictive models of protein modifications and phase separation from sequence and…
5 preprints · 1 published
Rank 5 Structural bio & protein design
Sequence rules and interfaces govern condensate architecture and phase behavior
Recent preprints and a published LLPS model connect sequence rules, interfacial transitions, and mesoscale control of condensate phase behavior.
5 preprints · 1 published
Rank 6 Learned protein and ligand representations for binding and perturbation predictionStructural bio & protein designRank 7 Computational recovery of sparse or unmeasured readouts in spatial transcriptomicsTranscriptomicsRank 8 Single-cell computational frameworks differ in how they shape analysis conclusionsSingle-cellRank 9 Latent generative models infer hidden single-cell states from count and spatial dataSingle-cellRank 10 Long-read and strain-level tools refine metagenomic genome recovery and classificationGenomicsRank 11 Sparse autoencoders and sequence models probe what biological foundation models encodeML / AI methodsRank 12 Computational tools strip contaminant and unwanted reads before downstream analysisTools & infrastructureRank 13 Learned sequence models locate short linear motifs and link disordered-region variants to…Tools & infrastructureRank 14 Simulations and cryoEM resolve conformational ensembles of protein, RNA and DNAStructural bio & protein designRank 15 Scalable software tools link variant calls to population and phylogenetic inferenceTools & infrastructureRank 16 Language model embeddings drive sequence-based prediction of protein and RNA functionML / AI methodsRank 17 Quantum-accurate energy models for molecular design and biomolecular force fieldsStructural bio & protein designRank 18 Deep generative and predictive models for molecules, peptides, and metabolite spectraML / AI methodsRank 19 Topological and morphological descriptors quantify structure in biological systemsStructural bio & protein designRank 20 Learned scoring and flexible alignment extend receptor-aware compound prioritizationML / AI methods
Show 20 more
Rank 21 Benchmarking and new callers target error modes in long-read variant detectionGenomicsRank 22 Closed-loop generation and structure-based scoring of designed biomoleculesStructural bio & protein designRank 23 Generative models sample biomolecular conformational ensembles and designable sequencesStructural bio & protein designRank 24 Confidence scores of learned structure predictors fail reference-based validationStructural bio & protein designRank 25 Linking noncoding sequence to gene regulation via chromatin, binding, and variant effectsGenomicsRank 26 Circuit topology and burst dynamics shape neural timescales and synaptic weightsSystems biologyRank 27 Portable software for genotype-based prediction and variant interpretationTools & infrastructureRank 28 Single-mechanism models reveal how isolated processes determine dynamical system fateSystems biologyRank 29 Deep learning and calibrated LD estimators for phenotype genes and positive selectionGenomicsRank 30 Dynamical models separate behavior-linked neural activity from internal computationSystems biologyRank 31 Simulation models project pathogen spread and gene-drive mosquito population suppressionSystems biologyRank 32 Genomic models link antimicrobial resistance genes to mobile elements and phenotypesGenomicsRank 33 Reference genomes, cell taxonomies and methylation maps in brain epigenomicsEpigenomicsRank 34 LLMs and knowledge graphs structure biomedical associations and ontology annotationML / AI methodsRank 35 Bounded LLM roles in explicit pipelines for single-cell metadata and cell annotationSingle-cellRank 36 Computational Tools Span Structure, Glycoform, and Chromatin AnalysisML / AI methodsRank 37 De novo and re-annotation methods expose hidden transposable element content in genomesGenomicsRank 38 Variable-length k-mers and genome composition scaling reframe sequence complexity analysisGenomicsRank 39 Phage antidefense proteins inhibit bacterial Thoeris and CBASS immune signalingGenomicsRank 40 Protocol dependence and attribution honesty in omics benchmarks and network controlSystems biology