TrendBio

Top 40

View all ›

BACKFILLED Code and industry signals are partial for backfilled days.

Rank 1

Single-cell

Computational methods integrate multimodal single-cell and spatial omics data

Recent preprints and papers add tensor, motif, and spatial-coordination methods for multimodal single-cell and spatial omics data.

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

ML / AI methods

Deep learning links histology, spatial, and single-cell omics with interpretable outputs

Foundation models and pathway-constrained networks now integrate proteomic, spatial, and imaging data with explicit interpretability.

5 preprints · 3 published · ↑ 2.9× vs 4-wk avg
Rank 3

Structural bio & protein design

Learned scoring of AlphaFold models and structural resources prioritizes binders

AlphaFold-based models now feed large-scale screening and interface scoring, making calibrated structural prioritization central.

5 preprints · 4 published · ↑ 2.8× vs 4-wk avg
Rank 4

Single-cell

Deep learning links single-cell, spatial, and microscopy data to phenotypes

Recent preprints and a published SEAGALL pipeline apply graph and transformer architectures to single-cell chromatin, spatial, and imaging data.

6 preprints · 1 published · ↑ 2.7× vs 4-wk avg
Rank 5

Structural bio & protein design

Simulated conformational and energetic states linked to functional and stability readouts

Head-to-head benchmarks of free energy and sampling methods now test simulated states against selectivity and stability readouts.

2 preprints · 6 published · ↑ 2.8× vs 4-wk avg
Rank 6 Learned latents steering protein conformational sampling and structure generation↑ 2.4×Structural bio & protein designRank 7 Pseudotime methods resolve gene trends and transient states in single-cell data↑ 2.2×Single-cellRank 8 Learned models now design RNA and codon sequences with structure-aware constraints↑ 2.3×Structural bio & protein designRank 9 Learned models predict RNA-protein binding, protein expression, and enzyme function↑ 2.1×ML / AI methodsRank 10 Benchmarks and verifiable FDR reporting test computational proteomics method reliability↑ 2.1×ProteomicsRank 11 Computational inference of microbial traits from genomic and expression data↑ 2.1×Systems biologyRank 12 Benchmarking ML models across splits, representations, and held-out targets↑ 2.0×Tools & infrastructureRank 13 Heterogeneous genomic resources reused for comparative and structural inference↑ 2.1×GenomicsRank 14 Benchmarking integrated omics and transcriptomic methods against ground-truth controls↑ 2.0×TranscriptomicsRank 15 Predictors quantify variant effects on dosage, stability, and protein-RNA interfaces↑ 2.0×GenomicsRank 16 Pangenome graphs and haplotype-resolved assemblies refine structural variant detection↑ 1.9×GenomicsRank 17 Deep learning on structured molecular inputs for property and interaction prediction↑ 1.8×ML / AI methodsRank 18 Encoding spatial and topological structure in high-dimensional biological data models↑ 1.9×TranscriptomicsRank 19 LLM and deep generative models mine bioactivity data and annotate metabolites from spectraMetabolomicsRank 20 Bayesian and ODE models infer molecular dynamics from single-cell and tracking dataEpigenomics