date="2019-10-01"
data_source="data_sources"
organism="Homo sapiens"

#assumes we're using samples.tsv and samples.tsv.header
study=$1
#human or mouse
org=$2
#sra, gtex, or tcga
dsource=$3
#RNA-seq, tissue, celltype Predictions/curations file
#e.g. Human.SRA_curated_predict.meta.w_rids.tsv
preds_file=$4

perl -e '$study="'$study'"; $study=~/(..)$/; $lo=$1; `mkdir -p $lo/$study`; print "$lo/$study\n";' > ${study}.dir
dir=`cat ${study}.dir`

cat samples.tsv.header <(fgrep "	$study	" samples.tsv) > $dir/samples.tsv.${study}
if [[ "$org" == "mouse" ]]; then
    organism="Mus musculus"
    date="2020-01-01"
    #sra specific MD file
    cut -f 1-43 $dir/samples.tsv.${study} | gzip > $dir/sra.sra.${study}.MD.gz
    #QC MD file
    cut -f 1-3,44-152 $dir/samples.tsv.${study} | gzip > $dir/sra.recount_qc.${study}.MD.gz
else
    #sra specific MD file
    cut -f 1-41 $dir/samples.tsv.${study} | gzip > $dir/sra.sra.${study}.MD.gz
    #QC MD file
    cut -f 1-3,42-150 $dir/samples.tsv.${study} | gzip > $dir/sra.recount_qc.${study}.MD.gz
fi

proj_header='rail_id	external_id	study	project	organism	file_source	metadata_source	date_processed'
dsource="$data_source/$dsource"
extra_proj_cols="$organism	$dsource	$dsource	$date"

#project MD file
cat <(echo "$proj_header") <(tail -n+2 $dir/samples.tsv.${study} | cut -f 1-3 | perl -ne 'chomp; $f=$_; ($rid,$run,$study)=split(/\t/,$f); print "$f\t$study\t'"$extra_proj_cols"'\n";') | gzip > $dir/sra.recount_project.${study}.MD.gz



#RNA-seq, tissue, celltype Predictions/curations
cat <(head -1 $preds_file) <(fgrep "	$study	" $preds_file) | gzip > $dir/sra.recount_pred.${study}.MD.gz

rm $dir/samples.tsv.${study}