Towards Estimating Gene Network using Structure Learning

S. Zainudin and S. Deris (Malaysia)


Bayesian Network, microarray, gene interaction, gene network.


Gene network is a representation of gene interactions. A gene usually collaborates with other genes in order to function. Past researches have successfully estimated gene network from gene expression microarray data. Gene expression data represent different levels of gene expressions for organisms during biological activity such as cell division. A framework for gene network estimation is to normalize gene expression data, discretise data, learn gene network and view network. Learning gene network uses hill-climbing approach that selects the highest scoring network in relation to the data as the best estimated network. This framework was used to learn the gene network for 3 subnetworks using S.cerevisiae gene expression data from Spellman dataset. The estimated subnetworks were compared to results from a past research. From 3 subnetworks, one subnetwork matched previous result, the other has mixed results and the last subnetwork did not match at all. This results show that Bayesian Network technique may be used to estimate gene network. However, better network may be estimated if previous knowledge is used with gene expression data.

Important Links:

Go Back