CTS

Longitudinal GWAS with CTS approach

Description

This R script run the longitudinal GWAS with Conditional Two-Step approach. Reference paper: https://www.nature.com/articles/ejhg20151#Abs1

Arguments

pheno_file: Full path to the phenotype file, longitudinal dataset with long-format

• RDS format
• covariates should be included for LMM analysis
• FID and IID columns should also be included
• Example pheno_file: /data/h_vmac/zhanm32/SPAREAD/DATA/ADNI/CTS_ADNI.rds

geno_file: Full path to the genotype file, file prefix included (PLINK .bim format)

id_col: Column name in the phenotype file representing subject IDs

time_col: Column name in the phenotype file representing time variable

y_col: Outcome variable column name

fixedEffects: Comma-separated list of fixed effects for LMM model after CTS (e.g., "Age,Sex,Education")

outpath: Full path to save the CTS outputs

outprefix: Prefix of output

Example Usaga

Rscript /data/h_vmac/zhanm32/SPAREAD/CODE/scripts/CTS_test.R
"/data/h_vmac/zhanm32/SPAREAD/DATA/ADNI/CTS_ADNI.rds"
"/data/h_vmac/zhanm32/SPAREAD/DATA/ADNI/ADNI_EUR_snpweights_baseline"
"ID"
"interval_years"
"SPARE_AD"
"Sex,PHC_Age_T1_MUSE,Education"
"/data/h_vmac/zhanm32/SPAREAD/OUTPUT/longitudinalGWAS/"
"EUR_ADNI"

Important Note

• Required packages: data.table, lme4, lmerTest, tidyverse, qqman
• Ensure that the phenotype file is cleaned before running this script:
  1. handle missingness yourself before running the R script
  2. make sure at least two time points per individuals
  3. make sure the pheno_file include FID and IID columns that matching the geno file
• Current code only support quantitative trait.