Top 40
View all ›BACKFILLED Code and industry signals are partial for backfilled days.
Transcriptomics
Graph and autoencoder models recover structure from sparse spatial transcriptomes
Recent methods pair learned representations with spatial, histology, or biological priors to handle incomplete spatial and single-cell transcriptome…
Single-cell
Ensemble and network frameworks infer biological state from multi-omics data
Benchmarking and interpretable models now quantify cell-type markers, subtypes, and biological age from single-cell and bulk omics.
ML / AI methods
Language models fused with structure, graphs, and dynamics for protein function
Protein language model embeddings are being combined with structure, dynamics, and generative components across function prediction and design papers.
ML / AI methods
Deep learning extends variant, regulatory and chromatin interaction prediction
Recent preprints and papers pair sequence-based deep models with orthogonal evidence such as MPRA, Hi-C and ChIP-seq to refine genomic predictions.
Single-cell
Benchmarked embedding methods are standardizing single-cell and spatial omics analysis
Spatial and single-cell omics now yield lineage, niche and 3D datasets that call for comparable, benchmarked representations.