Establishing large-scale gene regulatory networks using a gene-knowledge-embedded evolutionary computation method
With the increased availability of DNA microarray time-series data, establishing large-scale gene regulatory networks (GRNs) plays an important role in studying systems biology. S-system is well recognized as a suitable model to reconstruct the kinetic model of GRNs. However, the inference of an S-s...
Saved in:
Published in | 2011 IEEE International Conference on Computer Science and Automation Engineering Vol. 4; pp. 272 - 276 |
---|---|
Main Authors | , , , |
Format | Conference Proceeding |
Language | English |
Published |
IEEE
01.06.2011
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Summary: | With the increased availability of DNA microarray time-series data, establishing large-scale gene regulatory networks (GRNs) plays an important role in studying systems biology. S-system is well recognized as a suitable model to reconstruct the kinetic model of GRNs. However, the inference of an S-system model of N-gene genetic networks has 2N(N+1) parameters in a set of nonlinear differential equations to be optimized. The existing method iTEA has shown to be effective in reconstructing GRNs. In this work, an improved version of iTEA is proposed (named iTEA2) for efficiently establishing large-scale GRNs by incorporating the domain knowledge of gene regulation into the proposed evolutionary computation method. High performance of iTEA2 arises mainly from the novel method of encoding chromosomes in using the intelligent genetic algorithm. Let I be the maximal number of genes that directly regulated each gene. iTEA2 uses a hybrid encoding method which consists of regulation strength, gene number regulated, and binary control parameters in a chromosome. The value of the strength parameter is the kinetic order and control parameters indicate that the up- and down-regulated kinetic orders are active or not. The total number of parameters encoded in a chromosome for one gene is 5I+2, which is independent upon N. The experimental results show that iTEA2 is significantly better than iTEA in terms of accuracy, convergence speed, and robustness in establishing large-scale GRN. |
---|---|
ISBN: | 9781424487271 1424487277 |
DOI: | 10.1109/CSAE.2011.5952849 |