Class Presentation on November 8, 2002
By Parag Jain
My efforts toward the Project
- Delaunay Triangulation calculation for 4INS: I started working for the project by searching for an algorithm that would compute the Delaunay Triangulations for a molecule.
-Found out a divide and conquer algorithm that did the same.
-The divide-and-conquer algorithm subdivides the area into two partial areas, computes recursively the Delaunay triangulation of the partial areas and merges finally both triangulations.
- Distributed Molecular Surface (DMS): DMS calculates the molecular surface (SES) of a molecule. The molecular surface resembles the van der Waals surface of a molecule, except that crevices between atoms are smoothed over and interstices too small to accommodate the probe are eliminated. The surface includes cavities in the interior of the molecule, even if they are not accessible to a solvent molecule coming from the outside.
-Author: Conrad Huang, 2002.
-I used DMS to conduct several runs for a set of 16 PDBs.
-As of today the plot of the times that DMS took to compute the surface area versus the number of atoms in the molecule fits well with the graph of a second degree polynomial.
- Improved solvent excluded molecular surface area estimations using Boolean masks: This algorithm also calculates the SES for the molecule. However, this algorithm is based on the boolean mask approach that was proposed by LeGrand and Merz (Le Grand and Merz 1993). This algorithm takes the method of Le Grand and Merz one step further by using masks to calculate SES.
-Author: Christopher Bystroff, 2001.
-As of today I have used MASKER to conduct several runs for a set of 11 PDBs out of those 16 that I used for DMS.
-As of today the plot of the times that MASKER took to compute the surface area versus the number of atoms in the molecule fits well with the graph of a second degree polynomial.
Next slide