Adolescent depression symptoms GWAS

A repository of example scripts for REGENIE association testing in the elevated symptoms of depression in adolescence GWAS

LINK TO LIVE SOP: https://docs.google.com/document/d/1ppvOzH2Rjb-8Z2eqe86ct03E51MtS25q/edit?usp=sharing&ouid=108956179468703355713&rtpof=true&sd=true

Instructions

• Install REGENIE if not already installed: https://rgcgithub.github.io/regenie/install/
• To use example scripts as templates:
  • Clone this GitHub repository

git clone [email protected]:poppyzenzi/AD_GWAS_pipeline.git

• Edit scripts and set-up as appropriate

• If you already have a pipeline set-up or don't need templates, please still refer to our examples to make sure flags and thresholds correspond
• After running association testing, please make sure you have converted -log10P values to raw P values (see script convert-log10P.sh)
• Generate Manhattan and QQ plots using the 4.plot_results.R script provided
• Return back to the SOP document for how to upload and report your results

Contact

For any questions or uncertainties please contact Poppy Grimes at [email protected]

REGENIE flag options

These are also included in the shell scripts provided

Step 1

--bed: PLINK binary files prefix for .bed/.bim/.fam files

--extract: = List of SNPs that pass QC filter (We generate this list in PLINK, however if you have already performed sufficient QC on your genotyped data file, you can omit this flag)

--phenoFile: Path to your phenotype file with the header FID IID pheno_col

--phenoCol: Phenotype column name

--covarFile: Covariate file path with the header FID IID C1 C2 C3

--covarCol: Covariate column names

--bsize: Chunk size for analysis

--bt: Option for binary trait

--out: Path to output file

Step 2

--bgen: Path to imputed data. Here we use the .bgen format split by chromosome. See REGENIE documentation for more info on file input

--phenoFile: Path to your phenotype file with the header FID IID pheno_col

--phenoCol: Phenotype column name

--sample: Sample file corresponding to input bgen (can also use --keep / --remove flags to only include individuals that pass QC filtering, see REGENIE documentation)

--covarFile: Covariate file path with the header FID IID C1 C2 C3

--covarCol: Covariate column names

--bsize: Chunk size for analysis

--minMAC: Minimum minor allele count

--minINFO: Minimum imputation info score

--bt: Option for binary trait

--af-cc : To output A1FREQ in case/controls separately in the step 2 result file and N_CASES N_CONTROLS

--firth --approx: Firth LRT, computational speed up

--pThresh: P-value threshold to apply firth correction

--pred: Path to file containing predictions from STEP 1

--out: Path to output file(s)