Greetings,
I have been using modPipe for the past 2 weeks and I have three (two questions + one comment) specific questions to ask:
1) When I downloaded and installed modPipe, I connected it according to instructions to MODELLER and some other external tools such as BLAST. First question is about external programs. Can all ten external programs mentioned at the end of the introduction page at http://salilab.org/modpipe/doc/intro.html be installed or just the four external programs described with instructions at the page http://salilab.org/modpipe/doc/install-ext.html (MODELLER, BLAST, CD-HIT and PROCHECK) ?
2) My introduction to modPipe was through the demo example described at http://salilab.org/modpipe/doc/running.html. This example executed relatively quickly for default given settings in modpipe.conf. However, when I compare the modpipe.conf file with the modpipe.conf file given at http://salilab.org/modpipe/doc/modpipe.conf I see that the content of NRSEQDB tag is different. The NRSEQDB tag in demo/modpipe.conf points to demo/db/testdb.hdf5 (~4.5MB) while the NRSEQDB tag in http://salilab.org/modpipe/doc/modpipe.conf points to binary form of file uniprot90 (which is > 2GB if downloaded from http://salilab.org/modeller/downloads/uniprot90.gz). It is taking awfully long to get the results for a single protein in 1100 or 1000 hits mode if non-redundant sequence database is set to be the entire uniprot90 binary database file. Should the NRSEQDB tag point to uniprot90.hdf5 file obtained by compressing the Uniprot90 FASTA file, or is there some more compact and smaller representation for the non-redundant sequence database which could be used to get the results faster?
3) This question is partially in connection with question 2). I see that the NRSEQDB and some other databases given in modpipe.conf tags during execution are copied in local tmp/sequence_id/ directory for every protein sequence separately. In the case of request for large number of proteins to be processes this may quickly eat up disk space (unprot90 binary file is larger than 2GB).
Other than the above three issues, I think that I have setup modPipe correctly for execution.
Thank you and my best regards, ----------------- Dimitrije Jevremovic intern at Bioinformatics & Computational Biology Department