Date of Graduation
Master of Science in Statistics and Analytics (MS)
Second Committee Member
Third Committee Member
gene expression, gene regulatory network analysis, molecular signatures
Gene expression profiling by microarray has been used to uncover molecular variations in many different diseases. Complementary to conventional differential expression analysis, differential co-expression analysis can identify gene markers from the systematic and granular level. There are three aspects for differential co-expression network analysis, including the network global topological comparison, differential co-expression cluster identification, and differential co-expressed genes and gene pair identification. To date, most of the methods available still rely on Pearson’s correlation coefficient despite its nonlinear insensitivity.
Here we present an approach that is robust to nonlinearity by using the edge-count test for differential co-expression analysis. The performance of the new approach was tested with synthetic data and found to have significant results. For real data, we used a human cervical cancer data set prepared from 29 pairs of cervical tumor and matched normal tissue samples. Hierarchical cluster analysis resulted in the identification of clusters containing differentially co-expressed genes associated with the regulation of cervical cancer.
The proposed approach targets all different types of differential co-expression and it is sensitive to nonlinear relations. It is easy to implement and can be applied to any sequencing data to identify gene co-expression differences between multiple conditions.
Lin, A. G. (2019). Detecting Differentially Co-Expressed Gene Modules Via The Edge-Count Test. Theses and Dissertations Retrieved from https://scholarworks.uark.edu/etd/3475