genome_prepare/script/Swissprot.sh

#!/bin/bash
#set -e
type="proteins"
outdir="."
script_path=$(dirname `realpath $0`)
database_dir="/Business/psn_company/Work/Transcriptome/Datum/Public/Database/genome_prepare/Swissprot/uniprot_sprot.fasta.dmnd"
declare -A software db_path
software["CDS"]="diamond blastx"
software["proteins"]="diamond blastp"
usage(){
    echo -e "Usage:"
    echo -e "\t$0 -i cds.fa -o outdir -t CDS"
    echo -e "OPTIONS:"
	echo -e "\t-h: help information"
    echo -e "\t-i: fasta file"
    echo -e "\t-o: output dir;default:./"
    echo -e "\t-t: proteins or CDS default:proteins"
    exit 1
}

while getopts 'ht:i:o:' OPT;do
    case $OPT in
        h) usage;;
		t) type="$OPTARG";;
        i) fasta="$OPTARG";;
        o) outdir="$OPTARG";;
        ?) usage;;
    esac
done

if [[ ! $fasta ]]
then
	usage
fi


if [[ ! ${software[${type}]} ]]
then
    echo "ERROR:unknown -t $type" 
    exit 1
fi
mkdir -p $outdir

#cat <<eof
${software[${type}]} -q ${fasta} -d ${database_dir}  -o $outdir/blast.swissprot.txt -e 0.00001 \
--outfmt  6 qseqid sseqid pident length mismatch gapopen qstart qend sstart send evalue bitscore stitle \
-p 12 --more-sensitive --max-target-seqs 5
awk 'BEGIN{FS=OFS="\t"}{if(!a[$1]){print $1,$13;a[$1]=1}}' $outdir/blast.swissprot.txt > $outdir/Swissprot.txt
#eof