Abstract:
When designing novel drugs, the need arise to screen large databases of drug candidates (small synthesizable chemical structures) for structures that resemble active liga...Show MoreMetadata
Abstract:
When designing novel drugs, the need arise to screen large databases of drug candidates (small synthesizable chemical structures) for structures that resemble active ligands, i.e. small chemical structures that are known to react with the target protein. If several active ligands are known one might improve the quality of the search by taking all of these into account. This can be done by generating a meta-structure which summarizes the active ligands and use this meta-structure for querying the database. In this paper we propose a method for making such a meta-structure by making a multiple spatial alignment of a set of active ligands taking the flexibility of chemical bonds into account. We present two implementations of our method. One using Differential Evolution (DE) for numerical optimization, and one using the Nelder-Mead method for numerical optimization. We investigate the quality of the two implementations on a data set from a previous study in the field and conclude that the DE based implementation outperforms the NM based implementation.
Published in: 2009 International Joint Conference on Bioinformatics, Systems Biology and Intelligent Computing
Date of Conference: 03-05 August 2009
Date Added to IEEE Xplore: 25 September 2009
Print ISBN:978-0-7695-3739-9