Research:
Graduate studies:
Undergraduate studies:
Lab information:
Photographs
Contact:
Bioinformatics Group
School of
Computer Science
University of Waterloo
200 University Ave W
Waterloo, ON N2L 3G1
Canada
E-mail:
Dan Brown

|
Our research group designs, develops and assesses computational
tools for the exploration of genomic and proteomic data. As computer
scientists, we also study algorithmic questions inspired by and
related to biological problems. Finally, we collaborate with
biologists to study the usefulness in practice of the methods we
develop.
Featured Research Project
Structural Pattern Recognition
For the vast majority of proteins, similar sequence implies similar
function. This is the foundation for many tools in bioinformatics, and
has been instrumental in the annotation of newly sequenced genomes.
However, there are exceptions in which proteins with highly similar
sequence can have quite different biological functions, and conversely
proteins with very divergent sequences can be functionally similar. To
identify functional similarities independent of sequence, we have
developed structural pattern recognition techniques for identifying
binding and protein interaction surfaces. Antifreeze proteins provide
an example of proteins in which sequence is a poor predictor of
function. We have found that a common surface pattern (repeated carbon
atoms matching the ice lattice) is shared by antifreeze proteins,
independent of sequence or fold. This pattern can be used to predict
antifreeze activity, and has been used to identify a novel antifreeze
protein homologous to lipid transfer proteins. |
|