Hey guys,
Recently, I did some annotation of my de novo assembled transcriptome in a non-model plant RNA-seq project, using Sma3s.
I installed the softwares and modules required according to the documentation here.
Version:
Perl (v5.12.4),
BioPerl (1.006901, checked the version using the command perl -MBio::Root::Version -e 'print $Bio::Root::Version::VERSION,"\n"'),
CPAN (v1.9456),
local Blast (Blast 2.2.26, because Sma3s requires formatdb and blastclust, et al. scripts, so here is blast 2.2.26 rather than blast+),
the taxonomic division of UniProt database was plant division. I downloaded uniprot_sprot_plants.dat.gz and uniprot_trembl_plants.dat.gz, then uncompressed and used cat function to concatenate them to one file named uniprot_plants.dat for later use.
iMac, Mac OS X Lion 10.7.5, quad-core, 8 GB of memory.
The Sma3s.pl script can be found here at the bottom of the developer's site. The blast process went quite well, when it came to the annotation step, it appeared some error message like this below:
It first automatically generated some index files in the specific folder then ran blast process, these files contains:
I don't know what was wrong with my process, would you mind giving me a solution or tip?
Thank you.
Regards,
lzsph
1 answer
Try to comment out line 22 of the script (where the index type is hard coded) and then see what happens. I think what is going on is an index is created in the SDBM_File format, and then Bio::Index::Abstract is reading the index and expecting a dbm file from DB_File. Also, it looks like the script is expecting a fasta to index and a swissprot file of annotations to index, so make sure the script is getting the expected file formats (including the blast output). To comment further about what is going on, it would be necessary to inspect your files and run the script. These errors typically involve malformed data and/or something specific related to the Perl installation. Hopefully, that will help you diagnose the errors.
Log in to answer this question.
