Rigid Molecule Docking: FPGA Reconfiguration for Alternative Force Laws
- PDF / 474,820 Bytes
- 10 Pages / 600.03 x 792 pts Page_size
- 17 Downloads / 141 Views
Rigid Molecule Docking: FPGA Reconfiguration for Alternative Force Laws Tom VanCourt, Yongfeng Gu, Vikas Mundada, and Martin Herbordt Department of Electrical and Computer Engineering, Boston University, Boston, MA 02215, USA Received 1 May 2005; Revised 7 September 2005; Accepted 1 December 2005 Molecular docking is one of the primary computational methods used by pharmaceutical companies to try to reduce the cost of drug discovery. A common docking technique, used for low-resolution screening or as an intermediate step, performs a threedimensional correlation between two molecules to test for favorable interactions between them. We extend our previous work on FPGA-based docking accelerators, using reconfigurability for customization of the physical laws and geometric models that describe molecule interaction. Our approach, based on direct summation, allows straightforward combination of multiple forces and enables nonlinear force models; the latter, in particular, are incompatible with the transform-based techniques typically used. Our approach has the further advantage of supporting spatially oriented values in molecule models, as well as the detection of multiple positions representing favorable interactions. We report performance measurements on several different models of chemical behavior and show speedups of from 130× to 1100× over a PC. Copyright © 2006 Hindawi Publishing Corporation. All rights reserved.
1.
INTRODUCTION
Noncovalent bonding between molecules is basic to the processes of life and to the effectiveness of pharmaceuticals. Chemical experiments are not always practical for measuring binding strength, and may be prohibitively expensive for screening 100,000 or more drug candidates against one molecule of medical importance. Instead, a number of computational approaches have been developed. The most precise computational models are based on quantum mechanics, or on approximations such as density functional theory. These techniques are computationally exorbitant, however, and infeasible for answering the first question: at what approximate offsets and orientations could the molecules possibly interact at all? Less costly techniques are used for initial estimates of the docked pose, the relative offset and rotation that give the strongest interaction. Although most large biomolecules (substrates) and small-molecule drug candidates (ligands) can flex or rotate around chemical bonds, many applications [1–5] assume rigid structure as a simplifying approximation. This still allows the modeling of many different rules or force laws governing interaction between molecules, including electrostatic, geometric, atomic contact potential, solvent effect, and many others. Since its introduction, 3D correlation [3] has become a standard technique for determining the best fit between
digitized representations of rigid molecule approximations. The technique is based on 3D voxel grids representing the substrate and ligand. It uses correlation to detect strong similarity between the 3D structure of the ligand and the 3
Data Loading...
