TrendBio

BACKFILLED Code and industry signals are partial for backfilled days.

Rank 1

Transcriptomics

Models of RNA measurement structure recover cell types, isoforms and rare subtypes

Late-2025 preprints and publications extend RNA-seq modeling to rare subtype discovery, deconvolution, isoform resolution and spatial reconstruction.

9 preprints · 4 published · ↑ 3.2× vs 4-wk avg
Rank 2

Genomics

Lineage sampling bias in primers, metadata, and cohorts shapes microbial inference

Recent studies show that primer design, public metadata, and sampling frames each restrict which taxa are observed before inference begins.

2 preprints · 5 published · ↑ 2.7× vs 4-wk avg
Rank 3

Structural bio & protein design

Sequence-based deep learning predicts protein binding, interactions and allergenicity

Sequence-only and learned-embedding models now predict binding affinity, interactions and function, with growing attention to interpretability.

9 preprints · 2 published · ↑ 2.5× vs 4-wk avg
Rank 4

Transcriptomics

Omics tools embed statistical validation and benchmarking into analysis workflows

Recent preprints and a published package target error control, computational efficiency and reproducibility across regulatory, splicing, imaging and…

6 preprints · 1 published · ↑ 2.5× vs 4-wk avg
Rank 5

ML / AI methods

Deep learning models predict regulatory activity and 3D chromatin from DNA sequence

Pretrained sequence models are now paired with Hi-C data and standardized benchmarks to predict cell type-specific gene expression and chromatin organization.

3 preprints · 3 published · ↑ 2.4× vs 4-wk avg
Rank 6 Dynamic models infer cell-state transitions from temporally resolved single-cell data↑ 2.2×Single-cellRank 7 Graph neural networks and embeddings predict molecular and protein interactions↑ 2.1×ML / AI methodsRank 8 Gapless T2T Korean reference and population multiomic data extend genome resources↑ 1.9×GenomicsRank 9 Multi-omic metabolic models link regulatory circuits to cancer and gut disease↑ 1.8×Systems biologyRank 10 Generative and transport models recover cell dynamics and unmeasured gene expression↑ 1.7×Single-cellRank 11 Sequence and nucleosome determinants jointly shape chromatin structure and accessibility↑ 1.9×EpigenomicsRank 12 Deep learning models link genomic variants to expression, splicing, and mutation rates↑ 1.6×ML / AI methodsRank 13 Computational models predict toxicity, promiscuity, and kinase-substrate relationships↑ 1.6×Systems biologyRank 14 Attention models decode brain recordings for speech, emotion and language alignment↑ 1.6×ML / AI methodsRank 15 Protein language models guide antimicrobial peptide mining and sequence redesign↑ 1.6×Structural bio & protein designRank 16 Input selection and fusion strategies govern biomedical ML prognostic performance↑ 1.6×ML / AI methodsRank 17 Genomic data reveal domestication history, duplication dating and inherited methylation↑ 1.6×GenomicsRank 18 Adaptive tokenization and KV-cache compression extend genomic language model capabilities↑ 1.6×ML / AI methodsRank 19 Graph and deep learning models predict biological entity interactions and sitesML / AI methodsRank 20 Network methods prioritize gene modules, evaluated by causal and enrichment criteria↑ 1.6×Systems biology
Show 12 moreShow less

Sponsors