In article k6n at ixnews1.ix.netcom.com, fawhite at ix.netcom.com (Frank White) writes:
> Can anyone make reccomendations for programs to do restriction
> mapping and plasmid assembly that run on a SUN Sparstation? Our group has
> a Sparc 10 that they access via terminal programs over the network. We
> currently have some programs that were written in house several years ago
> but can't be updated with new enzymes etc, as well as being hard to learn.
In my spare time, I operate a facility at the University of Manitoba called
BIRCH (Biological Research Computer Hierarchy). BIRCH incoprorates a complete
suite of public-domain or freeware programs and databases running on networked Sparcstations.
Software packages that we currently have implemented include fasta, FSAP,
XYLEM, MBCRR, Barton's Oxford programs, Clustalv, GDE, PHYLIP, Zucker,
and various other programs. When GenBank, PIR etc. are taken into account,
we use about 800Meg. of disk space.
To get an idea of what the system looks like, download the BIRCH manual
in either PostScript (birch.ps) or WordPerfect (.wp) form from directory
'psgendb' at ftp.cc.umanitoba.ca. The manual over 2 years old now, but it
gives you a pretty good idea of how things run.
> The terminal programs can't handle X windows, so it's strictly command-line
> UNIX. We may switch to local software (Lasergene, Geneworks etc) in a year
> or so, but my boss is interested in seeing what's available now for the SUN.
> Any suggestions would be appreciated! Thanks!
I think you'd be making a big mistake to move back to the PC world when you
have Unix available. PC operating systems are moving towards increasing
complexity and will be harder and harder to administer as time goes on.
Things like networking, multitasking, security, virtual memory, 32-bit addressing,
and so forth are add-ons to PC-based operating systems, and as a result,
never seem to work very well. Furthermore, as hardware and software evolve,
you have to keep upgrading these things on EACH and EVERY PC.
A Unix server with X-windows clients solves most of these problems, and can
even be cost effective once you have a certain number of terminals. A Unix
system is not trivial to administer, but I'd wager that it is no harder to
administer a single Unix server than it is to try to manage multiple
independently-configured PC's. Networking, multitasking, security, virtual
memory etc. are intrinsic to Unix, and much of the pioneering
work was done in Unix. As I'm sure you know, most Unix implementations
on workstations are 32-bit implementations (although 64-bit addressing is
on the horizon for Sun's). Finally, as hardware and software evolve, you upgrade
one server and everybody instantly has the upgrade. On X-terminals,
about the only thing you really need to upgrade is memory, but beyond what
you need to manage windows on a the screen, there's no point in adding
While most programs available through BIRCH can be run on text-based terminals,
you really miss a lot by not using X-Windows. On our system, you can now run
almost everything through Steven Smith's Genetic Data Environment (GDE).
By accessing all of the programs & databases from a single front end,
GDE becomes a tremendously-powerful tool, that is at the same time easy to
use. Just as an example, as part of my Introductory Cytogenetics course,
I teach two lab sessions on computer analysis of sequences. In two 2hr.
sessions, the students learn enough of OpenWindows to get around, and then
learn how to use GDE. By the end of the second session, they are ready to
do a simple assignment requiring them to compare sequences for similarity,
print out the similar regions, and identify features from GenBank entries.
One of the strongest arguments against abandoning the PC world in favor
of Unix used to be that all the good commercial software runs on PC's.
That argument is no longer true. For example, I now have WordPerfect
and Lotus123 running on my Sparc. Things like dBaseIV, SAS, Mathematica,
CorelDraw, are all available for the Sparc environment. In addition to
commercial software, there is of course the massive collection of
software from the GNU project (compilers, editors, graphics). Finally,
much of the really exciting developments in molecular biology software
(eg. ENTREZ, GDE, ACEDB) were developed under Unix. (Let's also not
I haven't had a PC on my desk in 4 years, and I don't miss it. I can do
almost everything on my Sparc, and when I do have to go to a PC
(eg. to run AUTHORIN) I find it frustrating, because I always run into things
that would be so much more straightforward under OpenWindows.
I would even question the assumption that a Sparc server with X-terminals is
more expensive than the equivalent number of PC's. While the initial
hardware cost may be higher, if you have to keep upgrading PC hardware
every couple of years, and throw the PC's away after 4 or 5 years,
do you really save? (My 4 1/2 year old Sparc1+ can easily run the latest
SunOS release, and the latest software with no problems. A 4 1/2 year old
PC is practically a doorstop.)
Also, does it make more sense to 1) spend several thousand dollars buying
a commercial software package and get quarterly or annual database updates
on CD-rom, and THEN spend money every year for the latest release
2) Install the latest update of freeware by FTP the day it comes out,
and obtain database updates over the network as they come out -- all for free?
In the long run, I would argue that a Unix client-server system is probably
no more expensive and far more straightforward to administer than a
(hardly) comparable PC cluster.
Brian Fristensky |
Department of Plant Science | A question is like a knife that slices
University of Manitoba | through the stage backdrop and gives us
Winnipeg, MB R3T 2N2 CANADA | a look at what lies hidden behind.
frist at cc.umanitoba.ca |
Office phone: 204-474-6085 | Milan Kundera, THE UNBEARABLE LIGHTNESS
FAX: 204-261-5732 | OF BEING