Path: utzoo!utgpu!water!watmath!uunet!ig!daemon
From: SYEH@BIONET-20.ARPA
Newsgroups: bionet.molbio.proteins
Subject: Re: BUG IN FASTP
Message-ID: <6522@ig.ig.com>
Date: 1 Jun 88 17:41:31 GMT
Sender: daemon@presto.ig.com
Lines: 32

From: BIONET 

Dear Tim,

I have looked into this problem with XFASTP.  It appears that XFASTP
does not correctly swap high scoring sequences for low scoring
sequences when the maximum number of kept scores (1000 for the DEC
version) is reached.  Thus the histogram would have the correct score
record, if this is the case, and the list of scores would be missing
some top scoring sequences.

The version of FASTP which is accessed by FASTP-MAIL maintains a
larger array of scores, and I have not encountered the same problem
with it since I have not exceeded the maximum yet.
In addition, the FASTP-MAIL version is more recent and
may not have this program bug.

To correct this problem, we are working to install a FASTA mail
server.  As you are aware, FASTA is a newer version of FASTP, and we
are trying to focus our support on this version.

Sorry for the inconvenience that this has caused you.  This problem
with XFASTP (and XFASTN) has only come to our attention within the
last month.
 
Sincerely,

Spencer Yeh			ARPAnet: bionet@bionet-20.arpa
Scientific Consultant			      (for BIONET issues) or
BIONET					 SYeh@bionet-20.arpa
(415) 324-4363
-------