A New Software to Construct Gene Regulatory Networks From Microarrays Data

As basic building blocks of life, genes, as well as their products (proteins), do not work independently. Instead, in order for a cell to function properly, they interact with each other and form a complicated network. Recently, many researchers agree that most of biological ambiguous questions might be easily being answered if a sophisticated modeling of the gene regulatory network (GRN) was constructed. Also, GRN was used to help understanding disease ontology and reducing the cost of drug development. During the last decade, many GRN inference algorithms that are based on microarrays data have been developed to unravel the complexity of gene regulation. One of the major problems with microarrays is that a dataset consists of relatively few time points with respect to a large number of genes. The dimensionality and high degree of noise are interesting problems in GRN modeling. In this paper we proposed a new integrated algorithm to overcome these problems. Our software show good performance comparable to previous methods and many of produced edges have evidence in the literature data. The proposed method was applied to time series gene expression data of Saccharomyces Cerevisiae and could potentially be applied to other networks in yeast as well as higher organisms. 1