Top 40
View all ›BACKFILLED Code and industry signals are partial for backfilled days.
Rank 1
Transcriptomics
Computational models infer cell dynamics and spatial organization from transcriptomic data
Spatial transcriptomics and proteomics tools now pair explicit dynamical, perturbation, and spatial models with large-scale descriptive atlases.
12 preprints · 3 published · ↑ 3.8× vs 4-wk avg
Rank 2 Single-cell
Benchmarking and new algorithms for single-cell regulatory and expression inference
Single-cell and spatial transcriptomics now rely on benchmarked methods for clustering, feature selection, and transcription factor inference.
6 preprints · 8 published · ↑ 1.7× vs 4-wk avg
Rank 3 ML / AI methods
Few-label adaptation and stochastic training extend protein and genomic prediction
Few-label fitness prediction and leakage-free variant-effect modeling appear alongside benchmarks of genomic language models on RNA tasks.
4 preprints · 2 published · ↑ 1.8× vs 4-wk avg
Rank 4 Tools & infrastructure
Curated references and coverage-aware tools refine microbiome taxonomic profiling
Recent preprints and papers add reproducible references and profilers for fungal, cross-domain, and phage-host analysis of microbial communities.
6 preprints · 2 published · ↑ 1.6× vs 4-wk avg
Rank 5 ML / AI methods
Manifold embeddings and physics-informed models of neural and cognitive dynamics
Manifold embedding of dynamic brain data and physics-informed forecasting of Alzheimer's decline both target latent structure in neuroimaging.
3 preprints · 1 published · ↑ 1.6× vs 4-wk avg
Rank 6 Assumption-light variant and repeat inference from long and short sequencing reads↑ 1.5×GenomicsRank 7 Duplication history shapes noncoding regulatory elements and miRNA network position↑ 1.5×GenomicsRank 8 Open tools for managing, annotating, and visualizing derived omics and genomic datasetsTools & infrastructureRank 9 Multimodal and foundation models integrate single-cell data across contextsSingle-cellRank 10 Language models are being extended from single-cell representation to omics inferenceML / AI methodsRank 11 Algorithms infer cell identity, lineage, and marker genes from single-cell dataSingle-cellRank 12 Representation design for protein interaction and structure prediction modelsStructural bio & protein designRank 13 Structure-guided generative models now scaffold enzyme active sites de novoStructural bio & protein designRank 14 Mutation-level fitness and within-host dynamics inferred from pathogen sequence dataGenomicsRank 15 Association methods add shrinkage, functional annotation and multitrait structureGenomicsRank 16 Genetic circuits as tunable-threshold perceptrons and resource-coupled neural networksSystems biologyRank 17 Haplotype-aware linkage and crossover tools refine genetic association inferenceTools & infrastructureRank 18 Tractable parameter inference for stochastic and deterministic biological modelsSystems biologyRank 19 LLMs encode proteins, genomes, and DNA as tokens for prediction and molecular designML / AI methodsRank 20 Graph and invariant deep learning link tumor omics to clinical and molecular endpointsML / AI methods
Show 19 more
Rank 21 Genomic profiles guide cancer biomarker selection and treatment response predictionML / AI methodsRank 22 Coupled models link cell-level processes to cardiac valve and atrial mechanicsSystems biologyRank 23 Graph-based formulations resolve haplotypes under read ambiguity and mixed samplesGenomicsRank 24 Generative models produce annotated biomedical training data and impute missing valuesML / AI methodsRank 25 Telomere-to-telomere assemblies close Drosophila and Bactrocera genome gapsGenomicsRank 26 Computational methods infer protein localization and cell identity from proteomics dataProteomicsRank 27 Detection and size control reshape inferred kinetics in stochastic gene expressionSystems biologyRank 28 Ground-truth validation and quality control for genomic variant and structure callsGenomicsRank 29 Compressed indexes and collision-free hashing for scalable sequence analysisTools & infrastructureRank 30 Dual-encoder and convolutional models predict TCR and MHC peptide bindingML / AI methodsRank 31 Simulators and cross-engine checks are testing reproducibility of biology softwareTools & infrastructureRank 32 Metagenomics maps subglacial and Danish environmental microbiomes at scaleTools & infrastructureRank 33 Backbone-conditioned nucleic acid sequence design and RNA energy parameter inferenceStructural bio & protein designRank 34 Network models of brain dynamics are being tested against empirical neural recordingsSystems biologyRank 35 Generative and Bayesian samplers tested against protein conformational populationsStructural bio & protein designRank 36 Plasmids spread and diversify T6SS while sequence data sharpen taxonomic boundariesGenomicsRank 37 Rhythmic dynamics and higher-order information in neural and speech systemsSystems biologyRank 38 ML-driven platforms advance small-molecule identification and drug target profilingMetabolomicsRank 39 Simplified microbial community models require explicit criteria for when they failSystems biology