BG7 exits prematurely

Question

[mark.rose]

Hi All

I'm trying out some different bacteriall assembly annotation tools. Currently on bg7. However, when I execute it using the included test data it begins but exits prematurely without error. Any ideas how to fix this? Thanks

$ ../bin/bg7 -n escherichia_coli_o104_h4_str._ty-2482_1_supercontigs -g ./escherichia_coli_o104_h4_str._ty-2482_1_supercontigs.fasta -p ./EHEC_ReferenceProteins_17_08_2012.fasta -r ./EHEC_Reference_RNAs.frn -c ./genetic_code.txt   -G ./GenBankExternalData.xml -o bg7_out
logging your params to /home/rosema1/BioInfo/bin/BG7/bg7_example_input_files/bg7_out/params.log
it looks like this machine has 24 cores, will use that for blast settings
creating RNAs blast db
RNA blast db created: /home/rosema1/BioInfo/bin/BG7/bg7_example_input_files/bg7_out/escherichia_coli_o104_h4_str._ty-2482_1_supercontigs.rnas.db
if you want, you can check makeblastdb log: /home/rosema1/BioInfo/bin/BG7/bg7_example_input_files/bg7_out/escherichia_coli_o104_h4_str._ty-2482_1_supercontigs.rnas.db.log
running blastn: RNAs vs genome sequence
done! will do a basic results check now
checking if blast xml output file looks ok..

$ ls bg7_out/
escherichia_coli_o104_h4_str._ty-2482_1_supercontigs_RNA_blastn.xml  escherichia_coli_o104_h4_str._ty-2482_1_supercontigs.rnas.db.nin
escherichia_coli_o104_h4_str._ty-2482_1_supercontigs.rnas.db.log     escherichia_coli_o104_h4_str._ty-2482_1_supercontigs.rnas.db.nsq
escherichia_coli_o104_h4_str._ty-2482_1_supercontigs.rnas.db.nhr     params.log

Answer 1

[mark.rose]

OK, I figured it out. In the shell script that runs the execution you find the following code

echo "checking if blast xml output file looks ok..."
rnaBlastOk=$(tail -n 1 ${rnasVsContigsOutputPath} | grep -c '</BlastOutput>')
if [ ! "${rnaBlastOk}" = 1 ]; then
  echo "looks like the RNA vs contigs blast didn't work"
  exit 1
fi
echo "done! everything looks ok"

The problem with this is that it only looks for the BlastOutput string on the last line. Turns out that string occurs on the second to last line and the conditional does not catch the failed grep. So I changed it from '1' to '2' (here and in the protein blast later on) and now it runs.