Overview of scSCOPE Framework.
<p>(A) scSCOPE requires a gene expression matrix along with cluster annotations as input. (B) The input data undergoes Iterative Sparse Lasso Regression to identify genes capable of segregating two distinct groups. In each iteration of LASSO regression, data is split into training and testing...
Saved in:
| Main Author: | |
|---|---|
| Other Authors: | , , |
| Published: |
2025
|
| Subjects: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | <p>(A) scSCOPE requires a gene expression matrix along with cluster annotations as input. (B) The input data undergoes Iterative Sparse Lasso Regression to identify genes capable of segregating two distinct groups. In each iteration of LASSO regression, data is split into training and testing groups with the same cluster composition as in the original dataset. Only those genes consistently chosen in over a threshold (<i>θ</i>) of iterations are designated as Core Genes. (C) Core Genes are then subjected to Bootstrapped Co-expression analysis, where all the genes significantly co-expressed or differentially co-expressed with the core genes are identified as secondary genes. This analysis is run in a 60% sub-sample of the original data for 100 iterations and only stable gene interactions that appear in more than K iterations out of 100 are selected. (D) Each core gene, and its stable secondary genes identified through Co-expression analysis, then undergoes Pathway enrichment analysis. (E) Pathway analysis can identify not only pathways enriched in the cluster but also core-secondary gene interactions and their involvement across multiple pathways. (F, G) The results from co-expression analysis, differential expression, and pathway enrichment are analyzed together to determine marker genes and pathways for the cluster of interest. Each of these steps are repeated and run in parallel for each analysis.</p> |
|---|