paperKB
coga / coga-kb
Help
Sign in

Chunk #9 — Methods — Raw genotype data, imputation, and quality–control of target cohort

Source
Polygenic risk associated with post-traumatic stress disorder onset and severity.
Embedded
yes

Text

Blood samples were drawn at JJPVAMC or ISMMS and shipped to Emory University for SNP genotyping. The genotype data were obtained with the Infinium PsychArray BeadChip from Illumina (San Diego, CA, USA). Genotype calling was made with GenomeStudio. Samples were processed in two batches (owing to different sample arrival times). Per-sample genotyping rate, in each batch, was >99%, resulting in a total of 303,378 typed variants. Hence, no sample needed to be discarded due to quality–control (QC). This leaves 244 samples with genotype and phenotype data. Imputation was performed with standard steps. First, the genotype data were split into individual chromosomes. Then, strand orientation of genotyped data of each autosomal chromosome were checked and corrected with PLINK23. As pre-phasing was believed to improve imputation accuracy and speed, the study data were pre-phased with SHAPEIT24 with genetic map data for build 3725. Imputation was done for a window of 5 Mb at a time with IMPUTE226 using phased reference panel from 1000 Genome Project phase 3 data set. The imputed data were reassembled with GTOOL. The following criteria were used for