prokkaGFF2fcGTF.sh

#! /usr/local/bin/bash

# convert a gff file generated by PROKKA to a featureCounts()-compatible gtf
# 
# input file must have extension "gff"

if [ "$1" == "-h" ]; then
  echo $'\nConvert a gff file generated by PROKKA to a featureCounts()-compatible gtf\n'
  echo $'USAGE:'
  echo "`basename $0` [inputGFFfile.gff]"
  echo $'\nOUTPUT:\na GTF file with the same base name as the input'
  echo $'\nAUTHOR:\nangeloo@gmail.com'
  exit 0
fi

GFFIN=$1
GFFBASE=${GFFIN##*/}

# ${FOO/%from/to}
echo "working on ${GFFBASE}"
# gffread -T -F $GFFIN | sed -n '/exon/!p' | sed 's/transcript/exon/' > ${GFFBASE/%gff/gtf}

tmpfile1=$(mktemp)

gffread -T -F $GFFIN | grep "\tbarrnap:" | sed 's/$/ gene_biotype "ribosomal RNA";/' > ${tmpfile1}
echo ${tmpfile1}

tmpfile2=$(mktemp)
gffread -T -F $GFFIN | grep "\tAragorn:" | sed 's/$/ gene_biotype "tRNA";/' > ${tmpfile2}
echo ${tmpfile2}

tmpfile3=$(mktemp)
gffread -T -F $GFFIN | grep "\tProdigal:" | sed 's/$/ gene_biotype "protein_coding";/' > ${tmpfile3}
echo ${tmpfile3}

echo "merging tempfiles..."
cat ${tmpfile1} ${tmpfile2} ${tmpfile3} > ${GFFBASE/%gff/gtf}

rm -f ${tmpfile1} 
rm -f ${tmpfile2} 
rm -f ${tmpfile3}
echo "Done!"