We are in the final process of setting up in house Genbank database and Blast
search program.
I have downloaded both NCBI's blast and WU's blast and tested them on our
Linux and Digital Unix box. I only tested a few options which we are going to
use: Expect, Filter, NCBI-gi, Descriptions and Alignments. Wu have these
options also.
My question is: are there any siginificant differences between NCBI's
"blastall" and WU's "blast" program in terms of feature, performance (speed
and memory use)? What is your experience on both and do you have any
preference, and why?
here is my one-week experience:
1. I need to separately download and compile filter programs (such as dust)
for WU Blast.
2. I need to build "ans2fast" for NCBI Blast while WU Blast has "gb2fasta"
executable already available.
3. When I tried to run "formatdb" on a big Genbank file (~10GB). I always got
a core dump saying "Failed to allocate 6.2 mb" while I have 80GB available on
the disk in that directory.
If you reply, please send a copy to gmei at pavonis.com since I do not check
newsgroup often.
Thanks in advance.
Guang Mei
--------------------------------------
-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
