In-silico Drug Design and in-silico screening (Track)




A ‘reductionist’ approach in virtual activity profiling of drug molecules

Vibin Ramakrishnan
Rajiv Gandhi Center for Biotechnology, Thiruvananthapuram, Kerala – 695014, India

Abstract:

Computational methodologies have become a crucial component for many drug discovery programs, especially in the early stages of the discovery pipeline. Improved computational methods such as data mining have in some cases emerged as a cost effective supplement to screening huge chemical libraries. This requires canonicalization of 3D molecular structure to more computationally amenable notations in lesser dimension. We present a method to generate an intuitively accessible representation for ligand molecules, facilitating an efficient topology based similarity search. The concept christened as ‘clock model’, employs a ‘reductionist’ approach in converting the molecular structure to a 2D representation and then to a 1D fingerprint. The 1D fingerprint of ligand can be sequentially queried across a large database to screen molecules based on their potential pharmocophoric points and relative topology. We illustrate the efficiency and applicability of this approach using a database of 8200 ligand structures available in protein data bank.

Reference:

1. Aimy Sebastian, Andreas Bender and Vibin Ramakrishnan. Virtual Activity Profiling of Bioactive Molecules by 1D Fingerprinting Molecular Informatics (2010) 29, 773-779.