Naslov
Parallel Protein Docking Tool

Autori
Sović, Ivan ; Antulov-Fantulin, Nino ; Čanadi, Igor ; Piškorec, Matija ; Šikić, Mile

Vrsta, podvrsta i kategorija rada
Radovi u zbornicima skupova, cjeloviti rad (in extenso), znanstveni

Izvornik
33rd International Convention on Information and Communication Technology, Electronics and Microelectronics (MIPRO 2010) : Proceedings / - Opatija, 2010, 1333-1338

ISBN
978-953233050-2

Skup
33rd International Convention on Information and Communication Technology, Electronics and Microelectronics

Mjesto i datum
Opatija, Hrvatska, 24.05.2010. - 28.05.2010

Vrsta sudjelovanja
Predavanje

Vrsta recenzije
Međunarodna recenzija

Ključne riječi
MPI ; PDB ; protein docking ; spherical harmonics

Sažetak
Parallel Protein Docking Tool (PPDT) system is performing a shape based process of protein docking using spherical harmonics. The software is made from scratch in C++. Input data are two PDB (Protein Data Bank) files and desired docking parameters. The PDB files hold information about protein's atom coordinates in three dimensional space. Protein structures are transformed with spherical functions which allow fast rotation and translation of 3D shapes in space. By translation and rotation (with a given resolution of angles and distance) the surface complementarity is tested for each orientation of two proteins in 3D space. The orientations are then ranked according to the given scores. The output PDB file now holds orientations with best scores ready for visualization. As a part of the system a tool for visualization of molecules was developed. It's purpose is to provide a visualization of molecular structure in all phases of protein docking process. The visualizer also allows interactive translation and rotation of input proteins and evaluation of their complementarity in real time. Although docking tools that use spherical harmonics already exist, we implemented a parallel version using MPI (message passing interface). In this implementation any worker MPI process independently makes docking search on six dimensional subspace determined by parameters in its task. PPDT has high scalability, modularity, variable granularity of tasks and provides significant speedup comparing to a non-parallel version of software.

Izvorni jezik
Engleski

Znanstvena područja
Računarstvo

POVEZANOST RADA