regression.task.wdl

task regression {
    input {
        String phecode
        String phecode_directory
        String plink
        File covars        
    }
    
    File phecode_file = phecode_directory + phecode + ".tsv"
    
    File plink_bed = plink + ".bed"
    File plink_bim = plink + ".bim"
    File plink_fam = plink + ".fam"
    
    Int disk_size = ceil(size(plink_bed, "GB")*2 + size(plink_bim, "GB") + size(plink_fam, "GB"))*3 
    
    command {
        plink --bed '${plink_bed}' --bim '${plink_bim}' --fam '${plink_fam}' --pheno '{phecode_file}'               --covar '${covars}' --logistic --ci 0.95 --out '${phecode}'
    }
    output {
        File output_file = "${phecode}.assoc.logistic"
    }
    
    runtime {
        docker: "us.gcr.io/broad-dsp-gcr-public/terra-jupyter-aou:2.1.22" 
        memory: "7 GB"
        cpu: "2"
        disks: "local-disk " + disk_size + " HDD"  
        preemptible: 3
    }
}