nextflow.enable.dsl=2

process VCF2PROT {
  memory 2.GB

  container "${params.container_vcf2prot}"
  containerOptions "${params.containerOptions}"
  // echo true
  publishDir "${params.outdir}/vcf2prot", mode: 'copy'
  stageInMode 'copy'
  maxForks 1

  input:
    path vcf
    path vcf_filtered

  output:
    path "*.fasta"

  script:
    """
    #!/bin/bash
    workdir=`pwd`
    patient_name=\$(basename $vcf .vcf)

    grep '^#' $vcf > work1.vcf

    /opt/conda/envs/vcf2prot/bin/python3 -c "
import pickle
import numpy as np
import pandas as pd

with open('$vcf_filtered', 'rb') as fp:
    vcf_file = pickle.load(fp)
    
vcf_filtered=[]
for i in vcf_file:
    vcf_filtered.extend(i[1])
    
vcf_filtered = pd.DataFrame(vcf_filtered)

#drop duplicates, it returns error, consequence of overlapping genes on genome
vcf_filtered = vcf_filtered.drop_duplicates()

vcf_filtered = vcf_filtered.replace({'chr':''}, regex=True)

vcf_filtered = vcf_filtered.sort_values(by=[0,1])

with open('work1.vcf', 'a') as f:
    for line in np.array(vcf_filtered):
        li = str(line).replace('[', '').replace(']', '').replace('\\'', '').replace(' ', '\\t').replace('\\n', '')
        f.write(f'{li}\\n')
    " 

     
    awk '{gsub(/^##contig=<ID=chr/,"##contig=<ID="); print}' work1.vcf > \$workdir/work_2.vcf
    ###remove non canonical chromososms
    grep -v '[0-9]_\\|Y_\\|X_\\|Un_' \$workdir/work_2.vcf > \$workdir/work.vcf
    ###
    bcftools csq -f /home/omic/Homo_sapiens.GRCh38.dna.toplevel.fa -g /home/omic/Homo_sapiens.GRCh38.112.gff3 \$workdir/work.vcf --phase a --ncsq 120 -O v -o \$workdir/annotated.vcf
    /home/omic/vcf2prot/bins/Linux/vcf2prot -f \$workdir/annotated.vcf -r /home/omic/vcf2prot/MANE_transcipts_reference.fasta -v -g st -o \$workdir

    mv *.fasta \${patient_name}_transcript_id_mutations.fasta
    """
}