The surface of a molecule can be represented as a graph. We already have a program that reads a file containing a three-dimensional protein structure and writes out a graph representation of the surface in which the nodes of the graph represent atoms on the surface of a molecule and the arcs of the graph join pairs of adjacent surface atoms. In previous work this representation has been used to predict ligand binding sites on the surfaces lof proteins (Mehio et al., 2010). We now want to extend this work, and to use this graph representation of protein surfaces in various other ways.
|
|
Figure 1: Graph of adjacent atomic groups on the surface of a protein. |
There are many possible projects in this area, e.g.
Constructing database indexes to enable fast searching.
Use surface profiles to rank the docking orientations that are predicted by a macromolecular docking program.
Use this representation to study the shapes of antibody complementarity determining regions.
Use this representation to study interactions between helices in protein structures.
Use this representation to study packing between beta-sheets.
Develop profiles for classifying proteins and protein surface patches.
Mehio, W., Kemp, G.J.L., Taylor P. and Walkinshaw, M.D. (2010) Identification of protein binding surfaces using surface triplet propensities. Bioinformatics, 26, 2549-2555. doi:10.1093/bioinformatics/btq490
The course "Computational methods in bioinformatics" (Chalmers: TDA507, GU: DIT741) is recommended, but not required.