Please use this identifier to cite or link to this item: http://repositorio.unicamp.br/jspui/handle/REPOSIP/195724
Type: Artigo de periódico
Title: Packing Optimization For Automated Generation Of Complex System's Initial Configurations For Molecular Dynamics And Docking.
Author: Martínez, José Mario
Martínez, Leandro
Abstract: Molecular Dynamics is a powerful methodology for the comprehension at molecular level of many chemical and biochemical systems. The theories and techniques developed for structural and thermodynamic analyses are well established, and many software packages are available. However, designing starting configurations for dynamics can be cumbersome. Easily generated regular lattices can be used when simple liquids or mixtures are studied. However, for complex mixtures, polymer solutions or solid adsorbed liquids (for example) this approach is inefficient, and it turns out to be very hard to obtain an adequate coordinate file. In this article, the problem of obtaining an adequate initial configuration is treated as a packing problem and solved by an optimization procedure. The initial configuration is chosen in such a way that the minimum distance between atoms of different molecules is greater than a fixed tolerance. The optimization uses a well-known algorithm for box-constrained minimization. Applications are given for biomolecule solvation, many-component mixtures, and interfaces. This approach can reduce the work of designing starting configurations from days or weeks to few minutes or hours, in an automated fashion. Packing optimization is also shown to be a powerful methodology for space search in docking of small ligands to proteins. This is demonstrated by docking of the thyroid hormone to its nuclear receptor.
Rights: fechado
Identifier DOI: 10.1002/jcc.10216
Address: http://www.ncbi.nlm.nih.gov/pubmed/12692791
Date Issue: 2003
Appears in Collections:Artigos e Materiais de Revistas Científicas - Unicamp

Files in This Item:
File SizeFormat 
pmed_12692791.pdf270.42 kBAdobe PDFView/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.