Transcriptome data analysis
All gene expression data set with gene accession numbers GSE46315, GSE85391, and GSE13715 were obtained from gene expression omnibus (GEO) database (de Peppo et al., 2013; Gjorevski et al., 2016; Lai, Asthana, Cheng, & Kisaalita, 2011). For microarray data sets, in addition to the normalized data, all annotation files in the GPL formats were downloaded from the same database. Annotated gene expression data sets were uploaded into NetworkAnalyst web-tool (Xia, Gill, & Hancock, 2015; Zhou et al., 2019). For unnormalized data, log2 transformation was used to perform normalization and Limma algorithm was applied as a statistical algorithm to calculate differentially expressed genes (DEGs) (Xia et al., 2015; Zhou et al., 2019). As a general rule for all data sets, we used adjusted p-value <0.05 and log2 fold change>1 to identify DEGs. The list of highly affected genes with adjusted p-values <0.05 and log2 fold change >1 were used to draw Venn diagrams.
In addition to microarray data sets, we used a Mouse RNA-sequencing data set with accession number GSE85391. Read count matrix of this study was downloaded from GEO database (Gjorevski et al., 2016)and was analyzed with edgeR package in R to find differentially expressed genes (Robinson, McCarthy, & Smyth, 2010). DEGs were filtered using edgeR adjusted p-values <0.05 and log2 fold change>1.