Biology


University of Houston
One platform integrates biology education and research

Reproduced with permission from NeXT Computer, Inc.
A Reference Guide to NeXT in Higher Education, Fall 1992
ยช 1992 NeXT Computer, Inc


"I'm addicted to NeXT computers," says Dan Davison, assistant professor of biochemical and biophysical sciences at the University of Houston. "They're a powerful, cost-effective, easy-to-maintain UNIX machines. Plus, NeXT has the apps we need-LISP, Sybase, and Mathematica. In fact, in some situations, the software is more valuable to me than the hardware."

In his undergraduate- and graduate-level introductory courses in computational molecular biology, Davison uses NeXT computers with MidasPlus, the molecular modeling software created by scientists and programmers at the University of California, San Francisco.

"We start the course using Macs; when we begin doing more serious work, we get students on the NeXT machines, where we use MidasPlus and public domain sequence analysis software," explains Davison.

For the NeXT portion of the course, Davison assigns each student a different data structure from the Brookhaven Protein DataBank. Using MidasPlus, students display and manipulate the molecule, determine its molecular size (lens x width x size), mutate a residue, and then print before-and-after pictures. "With this exercise, I'm trying to get the students to see the changes a mutation can cause in a protein," explains Davison. "Students gain a better understanding of the molecules when they can look at them from different angles with a three-dimensional model."

Davison also uses NeXT machines in his research work. For example, using the Zilla application for distributed computation, one of Davison's graduate students, Ed Chen, has implemented a parallel distributed version of the standard genetic sequence library search algorithm. The algorithm parcels out searches to seven NeXT machines in Davison's department. "With Zilla, it takes two hours instead of all night for me to complete a search," he says. "We have about three-quarters of the ability of a supercomputer. In fact, overnight we can get the performance of about 1 Cray X/MP processor with Zilla and the NeXT machines."

Another project uses Lawrence Livermore Laboratory's genome project data dictionary and tables with Sybase on the NeXT machine to maintain the Chromosome 8 physical-mapping database, a collaboration between Biology Department faculty Dan Wells and Michael Wagner. "Sybase on the NeXT offers us high performance, ease of use, and reliability," says Davison.

He recently started using Interface Builder to create NeXTSTEP interfaces for programs he created on other platforms before obtaining a NeXT machine. "Interface Builder makes programming much more productive," he says. "I'm able to write fewer lines of code, and I have interfaces with a consistent functionality that are much friendlier to use."

For more information, please contact:

Dan Davison
Assistant Professor of Biochemical and Biophysical Sciences
University of Houston
4800 Calhoun
Houston, TX 77204-5934
(713) 743-8366
davison@uh.edu