Network Analysis of miRNA and Protein Expression Profiles in Breast Cancer Patients.

Zhang Yang,Cong Weihui,Wang Dehua,Zhao Yi

DOI: https://doi.org/10.1145/3121138.3121187

2017-01-01

Abstract:miRNAs and proteins are key regulators of gene expression, but their roles lead to tumor initiation and progression are not very well characterized. To better understand their dynamics in a systematic approach, this paper employs network analysis method to investigate their molecular activity and network-wide function in breast cancer patients. Random forest model has been used for filtering and wrapping miRNAs and proteins expression profiles from breast cancer patients' normal and tumor tissues. Maximal Information Coefficient of every two miRNAs or proteins is calculated as distance of two nodes in each network. Comparing the miRNAs network from normal tissue and tumor tissue, 10 nodes in two networks are selected respectively based on the differences in distance. Namely, hsa-mir-101-2, hsa-mir-10b, hsa-mir-130b, hsa-mir-190b, hsa-mir-193a, hsa-mir-204, hsa-mir-28, hsa-mir-365-2, hsa-mir-375, hsa-mir-192, including eight miRNAs are highly related with breast cancer. Similarly, E-Cadherin, PI3K, Caveolin, GSK3, XBP1, Bax, and syk have been identified in the proteins network associated with breast cancer. This novel method will yield deep insight into the regulatory architecture of breast cancer initiation and progression. © 2017 Association for Computing Machinery.