Genetic Algorithms for Targeted Drug Combination Design and Microarray Feature Selection with Cluster Analysis
Date Issued
2011
Date
2011
Author(s)
Lin, Han
Abstract
Combinatorial optimization is an important topic on finding an optimal set of objects from a finite set of objects in applied mathematics and theoretical computer science. Various problems in science, engineering, and bioinformatics may be described as within such category. In cases with large dimensionality, the exhaustive search method is often too slow to be of any practical usage. This thesis implements a Genetic Algorithm Programming Framework (GAPF) for solving combinatorial optimization problems with large dimensionality and with an uncertain or assigned amount of selection. GAPF is applied to practical bioinformatics problems including drug combination discovery and influential gene discovery. On microarray data analysis, the hierarchical cluster analysis of the resulting gene profile from GAPF shows a near optimal separation when compared with the original T-Test results. On targeted drug combination design, we designed a fitness function which considers multiple drugs against multiple genetic pathways with positive and side effects. GAPF results in solutions with minimum drugs usage and maximum pathway coverage.
Subjects
combinatorial optimization
feature selection
influential genes
drug combination design
anticancer drug
SDGs
Type
thesis
File(s)![Thumbnail Image]()
Loading...
Name
ntu-100-R97943161-1.pdf
Size
23.32 KB
Format
Adobe PDF
Checksum
(MD5):92fac91e41abbac85bcbd60c9650a80b