Link Search Menu Expand Document

Configuration

Before running nf-gwas, a Nextflow config file must be prepared. This file includes both pipeline parameters and cpus/memory directives. Additionally, the correct profile must be specified before executing the pipeline. Fully working examples including different parameters can be found here.

Examples

The nf-gwas pipeline currently supports imputed files in BGEN or VCF format (coming from Michigan Imputation Server). Please find below two working config files to run nf-gwas.

BGEN files

The following minimal configuration file runs an additive model using the UK Biobank data.

params {
  project                       = 'ukb-gwas'
  genotypes_array               = '/data/genotyped/ukb_cal_allChrs.{bim,bed,fam}'
  genotypes_imputed             = '/data/imputed/*bgen'
  genotypes_build               = 'hg19'
  genotypes_imputed_format      = 'bgen'
  phenotypes_filename           = 'phenotype/my_phenotypes.txt'
  phenotypes_columns            = 'phenotype1,phenotype2'
  phenotypes_binary_trait       = false
  regenie_test                  = 'additive'
  regenie_sample_file           = 'ukbXXX.sample'
  covariates_filename           = 'phenotype/my_covariates.txt'
  covariates_columns            = 'f.31.0.0,f.21022.0.0,f.22000.0.0,f.22009.0.1,f.22009.0.2'
}

process {

    withLabel: 'process_plink2' {
        cpus   =  4
        memory =  32.GB
    }

    //
    withLabel: 'required_memory_report' {
        memory =  32.GB
    }

    withName: 'REGENIE_STEP1|REGENIE_STEP2'
    {
        cpus   = 8
        memory = 16.GB
    }

}

VCF files

The following minimal configuration file runs an additive model using VCF data from e.g. Michigan Imputation Server.

params {
  project                       = 'mis-gwas'
  genotypes_array               = '/data/genotyped/allChrs.{bim,bed,fam}'
  genotypes_imputed             = '/data/imputed/vcfs/*vcf.gz'
  genotypes_build               = 'hg19'
  genotypes_imputed_format      = 'vcf'
  phenotypes_filename           = 'phenotype/my_phenotypes.txt'
  phenotypes_columns            = 'phenotype1, phenotype2'
  phenotypes_binary_trait       = false
  regenie_test                  = 'additive'
  prune_enabled                 = true
}

process {

    withLabel: 'process_plink2' {
        cpus   =  4
        memory =  32.GB
    }

    //
    withLabel: 'required_memory_report' {
        memory =  32.GB
    }

    withName: 'REGENIE_STEP1|REGENIE_STEP2'
    {
        cpus   = 8
        memory = 16.GB
    }

}

Table of contents