Publication Date

2012

Document Type

Honors Thesis

Department

Computer Science

Keywords

Computational biology, Software engineering, Macromolecules, Rigidity (Geometry), Biological assembly, Biological unit

Abstract

KINARI is a software development project undertaken by Ileana Streinu's Linkage Lab research group at Smith College and UMass Amherst. KINARI-Web, the current public release of the software, allows users to perform rigidity analysis on proteins and visualize the results [7]. Continuing e orts are being made to produce a more complete, easy-to-use, and reliable software. Protein Data Bank (PDB) les only contain the information that pertains to the asymmetric unit [3]. In many cases, especially for proteins solved via X-ray crystallography, the asymmetric unit is not the functional form of the protein, which is called the biological assembly or biological unit. In order to work with a protein in its functional form in KINARI, the coordinates and secondary structure information would have to be calculated a priori. Although there are tools that reason about biological assemblies, none are capable of being readily integrated with KINARI. Performing rigidty analysis on biological assemblies is especially pertinent to viral research. In most cases, virus capsids are made up of repeated protein subunits. The PDB entry includes only the unique repeating unit and the transformation matrices, which would produce the biological assembly when applied on the unique subunit. My main contribution is building the software tool KINARI BioAssembly. Cre- ating the software involved several steps, which included doing a proof of concept, coding the back-end, which takes the input PDB le and produces an output PDB le that has the biological assembly entry, coding the front-end, which is the graphi- cal user interface that detects for biological assemblies to be built and provides users i with options on how to generate and preview the biological assembly output, testing the software, both visually and using unit tests, and integrating BioAssembly into the KINARI pipeline. Presently, the integrated version of KINARI BioAssembly is available on the beta testing site, which is made accessible to the members of the Linkage Lab and students on the Smith and UMass campuses. The public version is set to be released by the end of May 2012 as part of KINARI-Web Version 1.3. Using KINARI BioAssembly, we were able to run benchmarking and pro ling tests on datasets containing various types of biological assemblies. The results of these tests show that (1) the KINARI software can handle proteins that are about 67% larger than the proteins that were previously tested, (2) there are grounds for making the rigidity analysis more e cient by taking advantage of biological assemblies that are composed of multiple copies of smaller subunits, and (3) the average run time for computing the bonds in biological assemblies grows at a slower rate than the average run time for computing the bonds in asymmetric units. Finally, using KINARI BioAssembly, we ran case studies on three types of virus related proteins: the aspartyl protease from HIV-1 isolate BRU (1HHP), the sca old- ing protein of the vaccinia virus (3SAQ), and the hexamer structure of the Rift Valley fever virus nucleoprotein (3OUO). All three case studies show that the rigidity of the biological assembly does vary from the rigidity of the asymmetric unit and that the degree of variance is di erent in each of the cases. The case studies also show that taking advantage of the symmetries in biological assemblies can improve the KINARI rigidity analysis algorithm. i

Language

English

Comments

x, 81 p. : col. ill. Honors project-Smith College, Northampton, Mass., 2012. Includes bibliographical references (p. 56-57)

Share

COinS