#20 · Single-cell
Learned multi-omics integration spans single-cell clustering and patient survival
The dominant thread is learned integration of heterogeneous omics views. scAMVIB applies an adaptive multi-view information bottleneck with maximum-entropy view weighting to multi-omics single-cell data, while scMUSCLE combines multi-subspace contrastive learning with structural smoothness and adaptive graph convolution for single-cell multi-omics clustering. CoFormerSurv extends this logic to patient-level outcomes, pairing an inter-omics Transformer with an inter-sample graph Transformer for multi-omics survival analysis. What is new is explicit modelling both across views and across samples, moving from cell-level clustering to clinical prediction. SGCRNA is a complementary Julia tool for co-expression network analysis of multi-omic data that avoids scale-free assumptions and tuning requirements. SDMap and So3D are resource papers rather than methods, providing spatial drug perturbation maps and 3D spatial transcriptome context with matched scRNA-seq references.
Published papers
Adaptive multi-view information bottleneck for multi-omics data clustering.
scAMVIB applies an adaptive multi-view information bottleneck clustering framework with maximum-entropy view weighting to integrate multi-omics single-cell…
Clustering single-cell multi-omics data via multi-subspace contrastive learning with structural smoothness.
Proposes scMUSCLE, a multi-subspace contrastive learning method with structural smoothness and adaptive graph convolution for clustering single-cell…
SGCRNA: spectral clustering-guided co-expression network analysis without scale-free constraints for multi-omic data.
SGCRNA is a Julia tool for co-expression network analysis that avoids WGCNA's scale-free assumption and tuning requirements, addressing slope neglect in…
SDMap: a comprehensive database of spatial drug perturbation maps.
The arrangement patterns and spatial distribution features of cells within tissues are key regulators of drug actions and cellular responses; however, the speci
So3D: a comprehensive three-dimensional spatial omics resource for decoding tissue architecture in physiology and disease.
So3D is a database and analysis resource integrating 72 3D spatial transcriptome datasets (882 slices, four species) with matched scRNA-seq references for 3D…
CoFormerSurv: Collaborative transformer for multi-omics survival analysis.
CoFormerSurv, a collaborative dual-Transformer for multi-omics survival analysis, pairs an inter-omics Transformer with an inter-sample graph Transformer to…
Topic history
| Date | Rank |
|---|---|
| #20 | |
| #23 | |
| #26 | |
| #10 | |
| #30 | |
| #26 | |
| #8 | |
| #27 | |
| #29 | |
| #33 | |
| #17 | |
| #37 | |
| #39 | |
| #9 | |
| #30 | |
| #29 | |
| #18 | |
| #26 | |
| #21 | |
| #24 | |
| #7 | |
| #1 | |
| #6 | |
| #4 | |
| #9 | |
| #12 | |
| #18 | |
| #34 |