TrendBio

Top 40

View all ›

BACKFILLED Code and industry signals are partial for backfilled days.

Rank 1

Transcriptomics

Learned models map spatial and single-cell omics to interpretable, predictive readouts

Recent preprints and papers couple spatial transcriptomics with histology and multiomic regulatory data in learned and statistical frameworks.

10 preprints · 8 published · ↑ 3.0× vs 4-wk avg
Rank 2

Tools & infrastructure

Reproducible long-read and amplicon software tools for microbial and fungal profiling

Benchmarked, containerized tools now cover long-read fungal, amplicon, and pathogen workflows with reported speed and error-rate gains.

6 preprints · 6 published · ↑ 2.1× vs 4-wk avg
Rank 3

Single-cell

Explicit baselines recover perturbation effects from pooled single-cell screens

Unmodeled baselines, sample structure, and measurement limits constrain which biological effects are recoverable from single-cell data.

6 preprints · 5 published · ↑ 2.0× vs 4-wk avg
Rank 4

Single-cell

Learned models recover latent structure from noisy and cross-modal transcriptomic data

Recent preprints and journal articles apply deep generative, graph and optimal transport models to sparse single-cell and spatial transcriptomic data.

2 preprints · 7 published · ↑ 1.7× vs 4-wk avg
Rank 5

Structural bio & protein design

Learned and knowledge-based models score protein-ligand and protein-protein interactions

Direction-aware statistical potentials and graph models for affinity and pose are appearing alongside QM/MD-fused and multi-task learners in recent preprints…

4 preprints · 5 published · ↑ 1.7× vs 4-wk avg
Rank 6 Learning reusable representations of count, noisy, and incomplete omics measurements↑ 1.7×ML / AI methodsRank 7 Protein and peptide language models guide controllable generative biomolecule design↑ 1.6×Structural bio & protein designRank 8 Context-aware modelling refines somatic variant detection in WGS, cfDNA, RNA and STR dataTools & infrastructureRank 9 New computational tools map splicing events, ac4C sites, and vesicle cargo effectsTools & infrastructureRank 10 Joint chromatin and sequence models resolve cell-type enhancer and regulatory syntaxSingle-cellRank 11 ML tools predict antibiotic action, pKa, carbohydrate binding and metabolic networksML / AI methodsRank 12 Reproducible tools span environments, visualization, workflows, and omics resultsTools & infrastructureRank 13 Held-out stress tests expose limited generalisation of biomedical predictive modelsSystems biologyRank 14 Diagnosing structure and sequence models for conformational and energetic fidelityStructural bio & protein designRank 15 Tokenized biological foundation models may distort continuous embedding geometryML / AI methodsRank 16 Image-based models now predict single-cell gene expression and whole-cell boundariesSingle-cellRank 17 Quantitative models of neural connectivity, dynamics, and manifold geometrySystems biologyRank 18 Experimental mutant data benchmarks ESM against SIFT and PlantCAD for missense effectsGenomicsRank 19 MSA perturbation and flow MD probe alternative protein conformationsStructural bio & protein designRank 20 Strain-level bacterial identification and quantification via MS proteomics pipelinesProteomics