Genome-Wide Pathway Approach to Dissecting Primary Open-Angle Glaucoma

Abstract

Purpose : Primary open angle glaucoma (POAG) accounts for approximately 75% of glaucoma in the Western world and is a phenotypically and genetically complex disease. Genetics play a significant role in POAG risk and more than a dozen loci have been identified that mediate risk for glaucoma and related endophenotypes. The NEIGHBORHOOD consortium has one of the largest collections worldwide of POAG cases and controls and one of our goals is to expand glaucoma genomics knowledge to better understand disease mechanisms and genetic architecture. In this study, we approached this goal by expanding the testable area of the genome beyond previous glaucoma genome-wide association studies (GWAS) through imputation, and by evaluating these results in the context of pathway analysis to elucidate potentially novel mechanisms of disease.

Methods : Single-variant results from the NEIGHBORHOOD meta-analysis were evaluated using the Pathway Analysis by Randomization Incorporating Structure program (PARIS) v2.4.0a4, a new Python implementation in the Biofilter 2.0 software suite using the Library of Knowledge Integration database, the CEU Region panel, 10,000 permutations, and the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway database. Because PARIS minimizes multiple testing and type I error concerns by implementing empirical genomic randomization to estimate significance, we set a threshold of P<0.0001; pathways attaining this threshold were interrogated further to identify significant (P<0.0001) genes.

Results : Of the 293 KEGG pathways, 33 were significant; 1882 unique genes comprised the significant pathways. Approximately 77% of these genes were only observed once, while 449 genes were significant in ≥1 significant pathway. Of the significant genes that appeared in significant pathways, 78% were only observed once, and 42 genes were in >2 significant pathways. Over 60% of the genes were individually significant in three pathways: selenocompound metabolism, phenylalanine, tyrosine, and tryptophan biosynthesis, and basal transcription factors.

Conclusions : Identifying novel POAG pathways will help to define the underlying mechanisms of POAG by identifying proteins and molecular pathways that influence pathogenesis. This information may help identify biomarkers for early molecular diagnosis and treatment and improve prevention strategies.

This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.