The proposed work will focus on improved 3-dimensional protein structure comparison and alignment. Local substructure recognition and local and global structure-structure alignment is by no means a solved problem and yet is vital to improving our ability to classify and understand protein structure en masse and for use in functional recognition. A problem of particular significance in the are of structural genomics where the number of functionally uncharacterized structures in increasingly rapidly. The starting point for this proposal is the Combinatorial Extension (CE) algorithm and associated database of pairwise alignments developed by the PI's of this proposal and in wide use by the community. There are three main interrelated aims of this proposal with Aim 1 having several sub-aims as follows:
Aim 1 - Improve the CE and MC-CE (Monte Carlo CE) pairwise- and multiple-structure comparison algorithms, respectively. This implies derivation of improved scoring functions for an alignment that:
Aim 1. 1-Interpret structure similarity/variability in a structural context- dependent (local versus global) manner Aim 1.2-Interpret structure similarity using a multiple rigid body approximation Aim 1.3-Interpret structure similarity in a scale-dependent manner to detect topological level similarities.
Aim 1. 4- Develop the concept of 'structural consensus' to build a multiple structural alignment.
Aim 2 - With the availability of improved structure alignments, characterized structures according to new and revised domain assignments and associated domain-level annotation.
Aim 3 -Provide all algorithms and associated annotated domain databases to a worldwide community in the same way CE and associated databases are made available today via a simple to use Web-friendly interface with associated graphical tools.
Aim 3 will draw heavily on the resources of the core, both computational and software.
Showing the most recent 10 out of 55 publications
