Towards Understanding the Relations Among Co-expressions, Annotations and Co-regulations

Raja Loganantharaj and Jun Chung



Microarray technology provides an opportunity to view transcriptions at genomic level under different conditions controlled by an experiment. From an array experiment, few hundreds of differentially expressed genes are selected and are clustered using one of several standard algorithms. Generally, coexpressed genes, which are members of the same cluster, are expected to have similar function, but unfortunately it is not often true. A functional annotation often refers to molecular function and biological processes and there is no specific way to combine the functional annotation when it comes to associating them with clusters. We will show that the proposed method based on singular value decomposition performs better than a typical smoothing technique. Co-regulated genes are often expressed simultaneously and are expected to have similar function. In this paper, we will propose a systematic method to study the relations among co-expression, co-regulation and functional annotations. We will illustrate the method with a recent microarray experimental conducted for identifying transcriptional targets of integrin a6b4 for breast cancer progression.

Index Terms Information content, gene ontology, annotation.

Full Text (PDF)