Quality Control Analysis of Genotyping Data in the Primary Open-Angle African American Glaucoma Genetics Study

Abstract

Purpose : The Primary Open-Angle African American Glaucoma Genetics (POAAGG) study is a 5-year project designed to elucidate the genetic architecture of primary open-angle glaucoma (POAG) in African Americans. A detailed quality control assessment of genotyping data from the discovery cohort (n=5500) was performed.

Methods : Genotyping was performed on the Illumina Infinium iselect platform on MEGAv2 (EX) Consortium Chip, consisting of approximately 2.04 million SNPs. Standard MEGA array content was supplemented with SNPs from other GWAS studies on POAG, several hundred additional mitochondrial variants, and variants detected from whole genome sequencing of POAAGG subjects. Sample success rate and call rates per sample and reproducibility of replicate pairs was analyzed using genome studio software. Gender mismatch issues, relatedness, sample quality, ancestry, and variant quality were assessed using PLINK.

Results : Successful genotypes from 1.87 million SNPs were obtained. An average call rate of 99.65% was achieved, with a sample success rate of 97%. The reproducibility rate between replicate pairs was 99.99%. PLINK annotated 73 samples as gender misannotations. Ancestry clustering by principal component analysis on study subjects, along with a set of 2.5 million 1KG samples, showed a homogenous African American population. No samples had a missing rate above 0.05. Relatedness in the study population was estimated using 434,685 polymorphic autosomal SNPs pruned for linkage disequilibrium. Unresolved relatedness pairs were excluded from further analysis. Variant quality filtering was applied sequentially based on missing call rate by SNP, minor allele frequency of 0 for all samples, Hardy Weinberg equilibrium p-value, and Mendelian error rate.

Conclusions : We successfully completed genotyping on the largest cohort of African Americans with POAG. Samples flagged at each quality filtering step were excluded from further analysis. Genome wide case control association studies based on single SNP analysis are currently underway.

This is an abstract that was submitted for the 2017 ARVO Annual Meeting, held in Baltimore, MD, May 7-11, 2017.