IUBio

BioLims --> BLAST

Andy Baxevanis andy at nchgr.nih.gov
Thu Jan 15 06:38:35 EST 1998


Tom Nelson wrote:

> What I really would like to hear are the opinions of labs who 
> are actually using the software, and not responses from software 
> marketing people.

Hi, Tom (and group!) --

We here at NHGRI have now been using the ABI BioLIMS software since the
summer; at the time we purchased the software, we were the only
non-commercial site that had done so, presumably because of the large
up-front cost involved.

While parts of the software work very well (specifically, the pieces
that allow the sequencers to communicate with the Sybase database), we
have, in our hands, found that the post-processing tools are, well,
lacking. (This is not to say that they would not be good for what other
people are doing; they just didn't do what *we* needed them to do.) 

What we ended up doing was to use the BioLIMS software to dump
everything into Sybase (which we were already extensively using anyway).
Once the data was in the database, we used a new set of programs that we
developed that automatically:

-- looks for new data in the database
-- screens for Alu and vector sequences
-- runs a variety of local BLAST searches on the masked data
-- generates HTML output presenting the significant hits,
   by way of a cgi application
-- allows for a monthly re-run of BLAST against the month
   database, to pick up any new hits
-- allows for commenting of individual sequences
-- supports multiple projects, with password protection
-- exports the data to Phred/Phrap for assembly

System requirements are Sybase (as required by BioLIMS), Perl 5.04
(freeware), and a Web server (we use Apache, also freeware).

Since the output is viewed through a Web browser, one of the advantages
is that sites do not need to purchase as many client licenses, since the
data can be viewed outside of BioLIMS proper. (These licenses are
separate from the instrument licenses, just to be clear.) We limit
access to the BioLIMS database itself to the people who actually need to
access it directly (for uploads/downloads and the like).

We've now used this package on two separate gene hunting projects and
the users, who have various degrees of facility with computers, have
generally found the tools to be very useful. We're currently in the
process of beta-testing, and hope to make a version public by the Genome
meeting in May. (Yes, public as in FREE, in the same way that NCBI makes
its software publicly available.)

If you have any questions, please feel free to ask. And good luck, Tom!

Andy



Andy Baxevanis, Ph.D.
Director, Computational Genomics
National Human Genome Research Institute
National Institutes of Health
Bethesda, Maryland 20892-4431




More information about the Autoseq mailing list

Send comments to us at biosci-help [At] net.bio.net