A collection of scripts to aid mixed linear model based GWAS with GEMMA

You need to have the following installed

• PLINK 1.90 ( https://www.cog-genomics.org/plink2/ )
• gemma (https://github.com/genetics-statistics/GEMMA )
• R (>=3.3))

• A binary PLINK dataset (.bed,.bim,.fam) and
• a phenotype file (with sample IDs in the first column and one or more phenotypes in subsequent columns). The header line should have unique names for phenotypes

Assuming one has a genotype dataset mydata.bed/bim/fam and a phenotype file pheno.txt with columns ID, Trait1, Trait2

1. Create necessary files for MLM GWAS (i.e. kinship matrix)

 ./compute_kinship.sh mydata

This creates a file result.cXX.txt in the directory ./kinship

Optionally, also compute principal components to be used as covariates

./makePCA.sh mydata

2. Prepare datasets for each trait

Rscript add-pheno-to-fam.R mydata.fam pheno.txt

3. Run GWAS

./RunGemma.sh mydata kinship/result.cXX.txt -p mydata.fam.Trait1.txt -t Trait1

This creates a directory output and places a file output/lmm-Trait1.assoc.txt (the GWAS result) as well as a log file.

There are additional options for RunGemma.sh

./RunGemma.sh plink_root kinship [ -c covariate ][-p phenotype ][-H phenotype has header][-t trait_name]

4. Plot results

 Rscript gwas-plots.R -t Trait1  -f output/lmm-Trait1.assoc.txt

Arguments:

• -f : GWAS output file
• -t : trait name to appear on plots