Characterizing Weighted Gene Correlation Network Analysis Module Quantitative Trait Loci Through Local and Distal Expression Quantitative Trait Loci Across Four Tissues in Diversity Outbred Mice

Zoe Vittum

Document Type

Article

Publication Date

Summer 2021

JAX Location

In: Student Reports, Summer 2021, The Jackson Laboratory

Abstract

Weighted gene correlation network analysis (WGCNA) is used to reduce the dimensionality of the transcriptome by clustering the tens of thousands of transcripts into a small number of clusters containing co-expressed transcripts. WGCNA groups transcripts based on correlation. Genes within a cluster are highly correlated to each other and often share functional enrichments. One approach to analyzing transcript clusters is to genetically map the first principal component of the transcripts in the cluster as a quantitative trait. The hope is that the mapped quantitative trait locus (QTL) will contain a master regulator gene that regulates the transcription of the other genes in the module, and that by identifying this regulator, we can link genotype to phenotype through functionally related groups of transcripts. If this is the case, we hypothesize that most module QTLs (mQTLs) will overlap numerous expression QTLs (eQTLs) of the transcripts in the module, and that these eQTLs will be heavily biased toward distal eQTLs, but include at least one local eQTL corresponding to the regulator of the distal eQTLs. Here we systematically tested that assumption. We found that there are at least four different characterizations of WGCNA mQTLs. The first characterization is the expected case. In this case explanatory eQTLs are dominated by distal eQTLs with few local eQTLs. The second class of mQTL was dominated by local eQTL with a few distal eQTL. The third class of mQTL was only explained by distal eQTLs, while the fourth class was only explained by local eQTLs. Overall, we found that the expected case discussed is not the rule for WGCNA modules, instead these modules represent a minority of the cases. Here we investigate all eQTL effect characterizations to help describe eQTL mapping patterns of mQTLs to aid in the explanation of how other mechanisms of gene expression are present within WGCNA mQTLs.

Please contact the Joan Staats Library for information regarding this document.

COinS