Fisher’s exact test was performed to check the enrichment of pMEIs in different regions of the affected genes. To test for enrichment in the eQTL analysis, common pMEIs were grouped into three categories based on their effect on gene expression: pMEIs not correlated with any gene (NS), correlated with at least one gene but not causal (related), and being causal for at least one gene (causal). For pMEIs grouped as NS and related, the affected gene of a pMEI is defined as the gene with the smallest FDR value by Matrix eQTL. For causal pMEIs, the affected gene is defined as the gene with a pMEI as the causal variant and with the smallest eQTL FDR value. pMEIs that are not within the 1-Mb window of any gene were excluded from the analysis. Functional genomic regions include enhancers from the Dragon Enhancers Database (DENdb, https://www.cbrc.kaust.edu.sa/dendb/src/enhancers.csv.zip) [56], 10 kb upstream of the transcription starting site (TSS), 10 kb downstream of the affected gene, and exons and introns of the affected gene. For each category, the number of pMEIs in different genomic functional groups was counted, and Fisher’s exact test was performed to determine the enrichment of pMEIs in those genomic regions in the related and causal categories relative to the NS category.
The enrichment analysis for sQTLs was performed similarly. For pMEIs grouped as NS and related, the affected ASE of a pMEI is defined as the ASE with the smallest FDR value; for pMEIs grouped as causal, the affected ASE is the ASE with a pMEI as the causal variant and with the smallest FDR value. The affected gene is the gene containing the affected ASE. If ASE includes transcripts from more than one gene, the longest gene among the overlapping genes was used to define the genomic functional groups. pMEIs that are not within the 1-Mb window of any ASE were excluded from the analysis.
Do you have any questions about this protocol?
Post your question to gather feedback from the community. We will also invite the authors of this article to respond.