TrendBio

Top 40

View all ›

BACKFILLED Code and industry signals are partial for backfilled days.

Rank 1

Tools & infrastructure

Benchmarked pipelines for metagenomic, long-read, and amplicon sequence analysis

Recent benchmarks expose hidden performance gaps in MAG pipelines while new Nextflow and GPU-accelerated tools target pathogen detection and amplicon data.

8 preprints · 5 published · ↑ 2.3× vs 4-wk avg
Rank 2

Single-cell

Models infer spatially resolved cell-cell interactions and localized gene expression

Spatial and single-cell datasets now support statistical models that separate interaction, localization, and regulatory inference at cell resolution.

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

ML / AI methods

Pretrained graph and sequence encoders coupled to task heads for molecular prediction

Recent work pairs pretrained or graph-based encoders with task-specific heads across protein, genomic, and glycan prediction tasks.

2 preprints · 9 published · ↑ 2.0× vs 4-wk avg
Rank 4

ML / AI methods

Discrete tokenization and sequence training strain embedding geometry in biology models

Embedding reliability and geometry, not benchmark accuracy alone, are now being tested across genomic, protein, and physics foundation models.

10 preprints · 1 published · ↑ 2.0× vs 4-wk avg
Rank 5

Tools & infrastructure

Somatic variant tools now model confounders, sample structure and germline background

Variant evidence is increasingly read from cfDNA fragments, bulk RNA and shared tumor genomes, where confounding context shapes interpretation.

5 preprints · 4 published · ↑ 1.7× vs 4-wk avg
Rank 6 Deep models recover structure from noisy, sparse and incomplete transcriptomic data↑ 1.6×Single-cellRank 7 Learned models link spatial transcriptomics, histology and single-cell RNA-seq analysis↑ 1.6×TranscriptomicsRank 8 Cross-cohort stress tests expose limited generalisation of biomedical predictive models↑ 1.6×ML / AI methodsRank 9 Computational tools and databases for RNA regulation, splicing and vesicle communicationTools & infrastructureRank 10 Reproducible environments, workflows and visual reports for omics analysisTools & infrastructureRank 11 Variant effect and calling methods benchmarked against experimental and population dataGenomicsRank 12 Integrative chromatin and multiomic methods link regulatory elements to target genesGenomicsRank 13 Mediation, FDR control, and regression models for high-dimensional omics dataGenomicsRank 14 Computational models probe conformational states and cross-family sequence reuseStructural bio & protein designRank 15 Computational tools predict molecular and metabolic properties from heterogeneous dataStructural bio & protein designRank 16 Protein language models predict viral, antibody and family-level protein propertiesML / AI methodsRank 17 Transformer and in-context models for neural, fMRI, and brain mesh dataML / AI methodsRank 18 Mathematical models link omics measurements to disease state transitionsSystems biologyRank 19 Genomic context and regulatory networks inform gene expression in yeast and BrassicaEpigenomicsRank 20 Intraspecies variability and pipeline choices govern bacterial proteomic accuracyProteomics