#!/usr/bin/env bash

# Stop on any error.
set -ueo pipefail

# This is where the unpacked reads will be stored.
mkdir -p reads

# Make a directory for indices and reference sequences.
mkdir -p refs

# Make a directory for runinfo.
mkdir -p info

# Set the project name.
PROJECT=PRJNA313294

# These will store the various SRR ids.
RUNINFO=info/zika.csv
SINGLE=info/single.csv
PAIRED=info/paired.csv

# Limit the read numbers to this.
READNUM=1000000

echo "*** Searching ENTREZ for project $PROJECT."
esearch -db sra -query $PROJECT | efetch -format runinfo > $RUNINFO

echo "*** Separating single end and paired end runs."
cat $RUNINFO | grep  SRR.*SINGLE | cut -f 1 -d , > $SINGLE
cat $RUNINFO | grep  SRR.*PAIRED | cut -f 1 -d , > $PAIRED

# This will run only if you have installed
# wonderdump that is necessary only for Bash on Windows
if hash wonderdump 2>/dev/null; then
    cat $SINGLE $PAIRED | xargs -n 1 echo wonderdump | bash
fi

echo "*** Unpacking single end reads."
cat $SINGLE  | xargs -n 1 fastq-dump -X $READNUM -O reads

echo "*** Unpacking paired end reads."
cat $PAIRED  | xargs -n 1 fastq-dump -X $READNUM --split-files -O reads

# This will be a file that comes with a prebuilt index.
IDX_FILE=refs/grch38/genome.1.ht2

# Check to see if we need to download the prebuilt index.
if [ ! -f $IDX_FILE ]; then

    # The URL to the hisat index.
    IDX_URL=ftp://ftp.ccb.jhu.edu/pub/infphilo/hisat2/data/grch38.tar.gz

    echo "*** Downloading the INDEX from: $IDX_URL"
    curl $IDX_URL| tar zxv

    # Move index directory to refs
    mv grch38 ./refs
else
    echo "*** Found HISAT index: $IDX_FILE"
fi


# This is the local GTF file.
GTF=refs/GRCh38.gtf

# Check to see wether we need to download the Gene Transfer file
if [ ! -f $GTF ]; then

    # The URL to the GTF file.
    GTF_URL=ftp://ftp.ensembl.org/pub/release-86/gtf/homo_sapiens/Homo_sapiens.GRCh38.86.gtf.gz

    echo "*** Downloading the GTF from: $GTF"
    curl $GTF_URL | gunzip -c > $GTF
else
    echo "*** Found GTF file: $GTF"
fi