Authors
Sanghamitra Bandyopadhyay, Santanu Santra, Ujjwal Maulik, Heinz Muehlenbein
Publication date
2007
Book
Analysis Of Biological Data: A Soft Computing Approach
Pages
183-201
Description
In this chapter we present an evolutionary approach for designing a ligand that can bind to the active site of a target protein. A tree like structure of the ligand is assumed, whose nodes are to be filled up from a library of functional units. Variable string length genetic algorithm (VGA) is used for evolving an appropriate arrangement of the basic functional units of the molecule to be designed. Since only the outer groups (i.e. those on the surface) take a major part in the van der Waals energy computation, a local search is first used to design the core part of the ligand by optimizing its internal energy. Once the geometry of the core part of the molecule is obtained, VGA is used for finding the appropriate arrangement for the outer layer of the ligand. The crossover andmutation operators are appropriately redesigned in order to tackle the concept of variable length chromosomes in VGA. Results are demonstrated for different …
Total citations
Scholar articles
S Bandyopadhyay, S Santra, U Maulik, H Muehlenbein - Analysis Of Biological Data: A Soft Computing …, 2007