APOL1 differential Co-expression analysis method and supporting data

Submitted by dkjang on

Tutorial for the Differential Co-expression View

Differential co-expressed gene modules were built using methods outlined by DiffCoEx. The network adjacency matrix is derived from differences in gene pair Spearman correlations between high-risk (HR) and low-risk (LR) groups. 1,273 genes were clustered into four modules: "Brown," "Purple," "Pink," "Green".

We compared hierarchical clustering of genes in each module in HR and LR separately to uncover the substructure within each module. This resulted in submodules labeled "A" to "O". We used Cytoscape (v3.7.2) to visualize the correlation networks utilizing the ExpressionCorrelation (v1.1.0) app (see figure on "View Data" tab).

The intra-module normalized connectivity was defined as the sum of adjacencies (from the matrix of adjacency differences) for each gene normalized by the largest connectivity value in the module. Thus, the most differentially co-expressed gene in each module would have normalized intra-module connectivity of 1. Hub genes were defined as genes with intra-module connectivity greater than 0.7.

Submodule eigengenes (1st principal component of module gene expression) were used to compare submodule gene expression patterns between HR and LR. The relationship between eigengenes can be explored through our interactive app.

Column descriptions

Gene Symbol: The gene symbol is hyperlinked to the gene page, which provides information for this gene from all analyses.

Gene Name: The descriptive name of the gene product.

Ensembl Gene ID: The Ensembl ID is hyperlinked to the gene page, which provides information for this gene from all analyses.

Module: The differentially co-expressed gene module identified by the DiffCoEx method ("Brown," "Purple," "Pink," "Green"); 'unassigned' genes were not clustered with differentially co-expressed genes.

Submodule: Submodule label; this identifies clusters of genes within each module that have different correlation patterns between HR and LR (see figure on “View data” tab).

Intra-module Normalized Connectivity: Connectivity of genes in the differential co-expression network; values range from 0 to 1. Larger values indicate stronger differential co-expression, i.e., the correlations between the gene and other genes within the module change between HR and LR samples.

research from
apol1_diff_co_exp