#! /bin/bash -l # note: this doesnt work from gg_job.pl because it doesn't use a parts.list of genome parts # fix somehow? or not SHARED=$HOME/scratch ACCOUNT=TG-xxx QTYPE=PBS #CLASS=NORMAL CLASS=dque CLOCKTIME=11:55:00 ncbibin=$HOME/bio/ncbi/bin aug=$HOME/bio/augustus # should be params: out_dir=`pwd` prot_dir=`pwd` proteins=$prot_dir/MY_GENES.aa protnam=queryaa faparts=1000 db_dir=$SHARED/ncbidb # use blast in parts on nrdb to save mem dbset="nr.00 nr.01 nr.02" #---------------- ## assume blast db is formatted. # for db1 in $dbset # { # if [ ! -f $db_dir/$db1.psq ]; then # $ncbibin/formatdb -p T -i $db_dir/$db1 # fi # } if [ ! -f $prot_dir/${protnam}1.fsa ]; then $aug/scripts/split_multifasta.pl --f=$protnam --seqs_per_file=$faparts \ --output_dir=$prot_dir \ --in $proteins fi qset=`/bin/ls -rt $prot_dir/${protnam}*.fsa` for qfile in $qset { for db1 in $dbset { query=`basename $qfile` jobn=bl$db1-$query if [ $QTYPE == "PBS" ]; then cat > $jobn < $jobn <