#!/bin/bash

set -euo pipefail

cd "$PROJECTHOME"/11.reference_v2

# HG sequences name should be changed.
cd raw
seqkit replace -k HG.map.txt -p "^(\S+).*$" -r "{kv}" GCA_030763125.1_ASM3076312v1_genomic.fna -o HG.fa -K --f-use-regexp

# MG & ZG use hapA genome
grep -Fwf MG.target.txt MG.gff > MG.target.gff
grep -Fwf ZG.target.txt ZG.gff > ZG.target.gff
cd ..

# Extract CDS sequences from GFF3 files using gffread
gffread -V -g raw/HN.fa -x HN.cds.fa -y HN.pep.fa raw/HN.gff3
gffread -V -g raw/HG.fa -x HG.cds.fa -y HG.pep.fa raw/Hippophae_gyantsensis.gff
gffread -V -g raw/HS.fa -x HS.cds.fa -y HS.pep.fa raw/HS.gff3
gffread -V -g raw/YN.genome.fa -x YN.cds.fa -y YN.pep.fa raw/YN.gff3
gffread -V -g raw/ZG.genome.fa -x ZG.cds.fa -y ZG.pep.fa raw/ZG.target.gff
gffread -V -g raw/ZJ.genome.fa -x ZJ.cds.fa -y ZJ.pep.fa raw/ZJ.gff
gffread -V -g raw/ZY.fa -x ZY.cds.fa -y ZY.pep.fa raw/ZY.gff3
gffread -V -g raw/Hippophae_tibetana.genomic.fasta -x HT.cds.fa -y HT.pep.fa raw/Hippophae_tibetana.mRNA.gff
gffread -V -g raw/MG.genome.fasta -x MG.cds.fa -y MG.pep.fa raw/MG.target.gff
cp raw/EA.gff3.cds EA.cds.fa
cp raw/EM.genome.cds.fasta EM.cds.fa
seqkit grep -s -v -r -p '[.*][A-Z]' raw/EA.gff3.pep > EA.pep.fa
seqkit grep -s -v -r -p '[.*][A-Z]' raw/EM.pep.fasta > EM.pep.fa

mkdir pep
mkdir cds
mv ./*.cds.fa cds
mv ./*.pep.fa pep

cd pep
# 保留主要转录本
primary_transcript ./ last_dot
# EM基因命名特殊性，导致直接使用primary_transcript会出现问题，直接复制即可
cp EM.pep.fa primary_transcripts/EM.pep.fa
# 移除终止密码子和序列描述
cd primary_transcripts
for i in *.pep.fa; do
    base=$(basename "$i" .pep.fa)
    echo "Processing $base"
    sed -i "s/\.$//g" "$i"
    sed -i "/^>/ s/[[:space:]].*//" "$i"
done