Multivariate canonical correlation analysis identifies additional genetic variants for chronic kidney disease

Area of Science:

• Genetics and Genomics
• Nephrology
• Statistical Bioinformatics

Background:

• Chronic kidney disease (CKD) has known genetic links to kidney function markers like estimated glomerular filtration rate (eGFR) and blood urea nitrogen (BUN).
• Previous genome-wide association studies (GWAS) using univariate methods have identified single nucleotide polymorphisms (SNPs) associated with eGFR and BUN.
• The potential for multivariate statistical methods to uncover additional genetic associations for kidney function remains largely unexplored.

Purpose of the Study:

• To investigate whether multivariate statistical analysis can identify novel SNPs associated with kidney function and CKD.
• To apply canonical correlation analysis (CCA) and meta-canonical correlation analysis (metaCCA) to discover new genetic associations.
• To validate the utility of multivariate approaches in genetic studies of kidney function and CKD.

Main Methods:

• Canonical Correlation Analysis (CCA) was applied to individual-level genotype data from CKD cohorts.
• Meta-Canonical Correlation Analysis (metaCCA) was used with published GWAS summary statistics.
• Identified SNPs were assessed for association with eGFR and BUN, and for colocalization with gene expression quantitative trait loci (eQTLs).

Main Results:

• The metaCCA method successfully replicated previously identified SNPs for kidney function, validating the approach.
• New lead SNPs associated with both eGFR and BUN were discovered, some showing functional impact predictions (e.g., in SLC14A2).
• Several novel SNPs were identified in European ancestry cohorts (CKDGen, NURTuRE-CKD, SKS), with rs3094060 showing significant association with CKD risk and FLOT1 gene expression.

Conclusions:

• Multivariate analysis using CCA significantly expands the discovery of genetic associations for kidney function and CKD beyond univariate GWAS.
• The identified novel SNPs and associated genes (e.g., SLC14A2, FLOT1) provide new avenues for understanding CKD pathogenesis.
• These findings highlight the value of advanced statistical methods in genetic epidemiology for prioritizing targets in CKD research.