Date: Tue, 05 Nov 1996 00:26:13 GMT Server: NCSA/1.5 Content-type: text/html Last-modified: Fri, 08 Sep 1995 19:34:50 GMT Content-length: 7381 Computational Biology in the UW-Madison CS Dept.

Computational Biology in the UW-Madison CS Dept

As a young science, computational biology offers a wealth of research opportunities. At the University of Wisconsin-Madison, scientists from chemistry, computer science, genetics, mathematics, molecular biology, plant pathology, and other disciplines are applying computational methods to various biological problems. Some investigations in the Department of Computer Sciences involve DNA sequencing and analysis, experiment management, modelling of ecological communities, protein-folding prediction, and species identification. Cross-disciplinary training programs are available for graduate students interested in careers in computational biology.

Research groups investigating computational problems in biology are in three subfields of computer science: artificial intelligence, databases, and theory. Each group collaborates with various biological laboratories on campus. A summary of some of the research activities follows.

The artificial intelligence group working with Professor Jude Shavlik applies machine learning techniques to several problems in molecular biology. Problems under investigation include: predicting protein secondary-structure; distinguishing protein-coding and noncoding regions; and recognizing promoters, splice junctions, terminators, introns, and ribosome-binding sites. Machine learning methods aid in the discovery of concepts underlying phenomena through the examination of multiple examples. For instance, a system can learn to find genes by examining many DNA sequences, each classified as to whether or not it contains a gene. This technique is powerful and potentially very valuable to the biological community. Currently, the primary research focus involves the incorporation of existing biological knowledge with computational discovery methods.

The database group headed by Professors Yannis Ioannidis and Miron Livny is developing a desktop experiment management system that will assist scientists in managing their experimental studies. The goal of the system is to have a single tool controlling the experimentation processes, managing the generated data, and efficiently processing user requests for data. An object-oriented database system is at the core of the system under development. This project proceeds in collaboration with several laboratories on campus, primarily those in the Departments of Soil Sciences, Molecular Biology, and Genetics. These groups are involved in simulation-based modelling of plant growth, microscopic imaging, and DNA sequencing, respectively.

The research group led by Professor Deborah Joseph applies techniques from theoretical computer science to develop algorithms for computational biology applications. One research project in collaboration with the Wisconsin E. coli Genome Project is leading to computational methods that generate accurate alignments of overlapping DNA sequences. Other work analyzes sequence data for interesting biological features. For instance, in collaboration with a group in plant pathology, one class of DNA sequence is being used to develop quantitative methods for identifying possible biological control organisms in ecological communities. Some of the algorithms developed by this project have been implemented on the department's parallel computers.

Two training programs offer cross-disciplinary training for doctoral students early in their graduate programs. The NIH-funded Biotechnology Training Program was established to train scientists and engineers to effectively apply interdisciplinary research tools to solve problems of biotechnological significance. This program has students throughout the biological and physical sciences involved in research problems of specific biotechnological relevance. The Applied Mathematics Training Program is being established to train scientists to effectively apply mathematical and computational tools to a wide range of scientific endeavors. Although specific to mathematical applications, students in this program can address a broad range of research problems including many in the biological sciences. Traineeships in both programs can be awarded to students entering graduate school.

Training Programs

Amy Kryder (kryder@cs.wisc.edu) and Carolyn Allex (allex@cs.wisc.edu) are current holders of Biotechnology Training fellowships; feel free to contact them with questions about the program.

Graduate Study in Computer Sciences

Electronic Access to Wisconsin Papers

The Wisconsin CS department maintains an electronic (ftp) archive of technical reports, other papers, and software.

The subdirectory machine-learning/shavlik-group contains additional papers by Shavlik's research group. See the file abstracts for a list of papers, which are in compressed postscript. (The papers in the files shavlik.tr92.ps and craven.mlrgwp93.ps are recommended as the first ones to read.)

Some Interesting Links


Last Changed: February 22, 1995 by shavlik@cs.wisc.edu