Token recognition error

[zillurbmb51]

Hi, I have a process as below:

process merging {
publishDir params.out, mode:'copy'
input:
tuple val(chrom),path(bgens) from imputed.map{it}
tuple val(chrom_mg),path(beds) from impute_merge.map{it}.first()
val flag from done_ch
output:
tuple val("${chrom}"),path("${chrom}.*") into merged
script:
def (bed,bgen,bim,fam,gz,sample,stats) = bgens
def (bed_mg,bgen_mg,bim_mg,fam_mg,gz_mg,sample_mg,stats_mg) = beds
"""
echo "${fam}" | sed "1d" | sed "s\/....\$//" > "${chrom}.merge-list.txt"
plink --bfile "${bed_mg}" "${bim_mg}" "${fam_mg}" --merge-list "${chrom}.merge-list.txt" --make-bed --out "${chrom}.step10.imputed.merged  || true
"""
}

I am getting error: Script compilation error: cause: Token regnition error at: ' echo "${fam}" | sed "1d" | sed "s\/....\$//" > "${chrom}.merge-list.txt" Details attached. Not sure what I am doing wrong. First time I am getting this error.

[pditommaso]

This looks a problem escaping the $ in the sed oneliner. Better to convert it into a separate Bash script and add into the pipeline bin/ folder https://www.nextflow.io/docs/latest/sharing.html#third-party-scripts