IUBio

Molecular Graphics on PCs

CZJ at CU.NIH.GOV CZJ at CU.NIH.GOV
Tue Jul 21 15:08:09 EST 1992


Reinhard Doelz writes
> Sorry, but that sounds a bit dangerous. The term molecular graphics implies
> the use of graphics. GRAPHICS. Depending on what you want to achieve, I find
> that 8 bits are not suitable for usual molecular modelling. Even if so,
> you will need to have a better than VGA graphics anyhow. If you ever saw
> a good 3D representation of a molecule on a PC, I doubt that it was a protein.
> You might have looked at a peptide instead. Simple rotations get you the
> imagination of depth, but if it takes a couple of seconds to redraw the
> screen you can't call this 'animation' any longer.
>
> Talk about prices. If you purchase a reasonable 486 PC with, lets say, 450
> MB hard disk, streamer, CDROM , large screen, mouse, operating system
> and draft printer it will cost you about $ 8-10 K in Europe (if you buy
> clones). For twice the price you get a damned good workstation. If you
> buy a cheap version (i.e., small disk, no streamer, etc) you get down to
> about $15 K. However, this workstation has power to display molecules rather
> than throw them on-screen occasionally.
>
> Talk about performance. The MFlops are getting cheaper these days, and
> a workstation of today is old-fashioned tomorrow. So why buy a chip which is
> expected to be no longer to of the crop next year ? The successor is currently
> named P5 and will be shown in equipment at the comdex exhibition Fall this
> year.
>
> Concluding, I have a Indigo on my desktop and there is also a Mac.
> I have not invested in PCs because of price/performance on the LONG run.
> I think it's better to go the workstation road if you want to do
> modelling.
>
> --

I would not be so pessimistic.  I just bought a 486 33 MHz clone
without monitor or printer but with 4 Mbytes of memory, 125 Mbyte
hard disk, and CD-ROM drive for $1,800 stateside from my local
clone bank.  Addition of a 14 in high resolution monitor would
have added $ 300 to the price.I have run PCMODEL (as noted in my
previous message) and it will take a protein structure allow you
to zoom in, rotate, etc.  On a 486 this is fast.  The problem
with the program is that documentation is skimpy, an essential
file conversion program is missing, no support or maintenance.

Without koncking MACs or other workstations, PC's have finally
become real computers.  Missing is high quality molecular graphics software
at reasonable prices.

Jim Cassatt





More information about the Bio-soft mailing list

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