Co-sponsored by The New Jersey Commission on Science and Technology Initiative on Structural Genomics and Bioinformatics and by Rutgers Bioinformatics Initiative.
This workshop will investigate models of protein classification and
protein interaction. It seems both impossible and unnecessary to
describe in detail every unique protein. To understand how a
protein's machinery works in a cell, one wants to develop an efficient
classification of proteins. It is well accepted that such a
classification should be based on particular pieces of proteins which
are called domains. Unfortunately, there don't exist precise formal
definitions of domains which are accepted by most experts, either for
amino acid sequence or 3D-structure presentations of proteins. There
are a few databases which present classifications and one can consider
them to infer experimentally presented definitions. For instance, the
SCOP database or FSSP are such databases for protein domains formed by
structural data and PRODOM and PFAM are protein domain databases
organized around sequence information. Acceptable formal definitions
are important for genetic engineering, drug design, and other fields
of modern biotechnology, because only through them can one build
software to manipulate proteins in silico, and thus we will explore
them in this workshop.
Comparison analysis is a fundamental part of many protein studies. We
will pay particular attention to similarity measures for proteins in
different representations. There are a lot of tools for such
analysis, but we don't yet know the limits of their usefulness.
Moreover, we know very litle about how existing tools can work
together in some integrated way. We will discuss these questions in
the workshop.
Next: Call for Participation
Workshop Index
DIMACS Homepage