paperKB
coga / coga-kb
Processing
Help
Sign in

Chunk #11 — Materials and methods — PRS features

Source
Predicting alcohol use disorder remission: a longitudinal multimodal multi-featured machine learning approach.
Embedded
yes

Text

PRS based on GWAS weights from 47 phenotypes were derived from 12 publicly available large-scale GWAS of alcohol-related traits conducted in EA and AA males and females including GWAS of alcohol consumption27–29, DSM-IV alcohol dependence28,30,31, and a maximum number of alcoholic drinks within 24 h. Additional PRS were derived from GWAS of other traits known to correlate with alcohol use and problems, including educational attainment32,33, anxiety disorders34, personality traits(e.g., aggression35, neuroticism33,36), depression33, subjective wellbeing32, brain structure37, and environmental sensitivity38 (overall number of PRS features = 1162). Details regarding the discovery of GWAS, including the number of individuals who participated in the GWAS and phenotypes, can be found in Table S6. Information on genotyping and quality control is available in the Supplemental Materials. Briefly, the well-established process of clumping and thresholding was used39 where single nucleotide polymorphisms (SNPs) from discovery GWAS were clumped based on linkage disequilibrium (LD) in the 1000 genomes EUR panel using PLINK 1.940, based on an R2 = 0.25, with a 500 kb window. SNPs were weighted using the negative log of the association p values. Scores