Differential correlation network analysis of rectal transcriptomes reveals cystic fibrosis-related disturbance.

ABSTRACT

Background: Intestinal pathology in cystic fibrosis (CF) remains mechanistically underexplored. Aim: We hypothesized that differential correlation network analysis of expression data would reveal hub genes of CF-related disturbance in the large bowel. Materials & methods: Transcriptomes of 29 rectal tissue samples were accessed at ArrayExpress (E-GEOD-15568 by Stanke et al.). Results: We identified 279 transcript pairs differentially correlating in CF and controls, including ESRRA and RPL18 (rCF = 0.55; rcontrols = -0.68; padj = 1.60 × 10-100), SRP14 and FAU (rCF = -0.69; rcontrols = 0.48; padj = 2.99 × 10-90), SRP14 and GDI2 (rCF = -0.34; rcontrols = 0.60; padj = 1.05 × 10-78). The genes related to membrane protein targeting (q = 8.34 × 10-14) and one cluster was clearly linked to male infertility. Conclusion:FAU, SRP14 and GDI2 may be involved in a compensatory protein trafficking mechanism in CF rectum, highlighting their potential therapeutic value.