Top 40
View all ›BACKFILLED Code and industry signals are partial for backfilled days.
ML / AI methods
Learned embeddings and linear models predict perturbation and binding outcomes
Recent preprints apply self-supervised, contrastive, and retrieval-based learning to transcriptomic perturbation and genomic prediction tasks.
Genomics
Computational methods map non-coding variants and TF activity to gene regulation
Recent preprints and publications apply DNA language models, single-cell multiomics, and network methods to dissect regulatory gene control.
Transcriptomics
Deep learning fuses histology with spatial omics for expression and deconvolution
Large-scale pretraining and cross-modal alignment now enable histology-based inference of spatial gene expression and spot deconvolution.
ML / AI methods
Graph and foundation models encode gene networks for perturbation and spatial tasks
Recent preprints couple GO, coexpression, and pathway graphs with deep models for perturbation, spatial, and kinase inference.
Single-cell
Single-cell models infer regulatory networks and cell states across omic modalities
Recent work pairs sequence-aware deep learning with single-cell multiomic and transcriptomic data to resolve regulatory architecture at cell-type resolution.