Identification of key regulators in prostate cancer from gene expression datasets of patients

Irengbam Rocky Mangangcha,Shakir Ali,R. K. Brojen Singh,Ömer Küçük,Md. Zubbair Malik

doi:10.1038/s41598-019-52896-x

Irengbam Rocky Mangangcha, Shakir Ali + Show 3 more

Open Access

https://doi.org/10.1038/s41598-019-52896-x

Copy DOI

Abstract

Identification of key regulators and regulatory pathways is an important step in the discovery of genes involved in cancer. Here, we propose a method to identify key regulators in prostate cancer (PCa) from a network constructed from gene expression datasets of PCa patients. Overexpressed genes were identified using BioXpress, having a mutational status according to COSMIC, followed by the construction of PCa Interactome network using the curated genes. The topological parameters of the network exhibited power law nature indicating hierarchical scale-free properties and five levels of organization. Highest degree hubs (k ≥ 65) were selected from the PCa network, traced, and 19 of them was identified as novel key regulators, as they participated at all network levels serving as backbone. Of the 19 hubs, some have been reported in literature to be associated with PCa and other cancers. Based on participation coefficient values most of these are connector or kinless hubs suggesting significant roles in modular linkage. The observation of non-monotonicity in the rich club formation suggested the importance of intermediate hubs in network integration, and they may play crucial roles in network stabilization. The network was self-organized as evident from fractal nature in topological parameters of it and lacked a central control mechanism.

Highlights

Identification of key regulators and regulatory pathways is an important step in the discovery of genes involved in cancer
The application of network theoretical methods to the protein–protein interaction (PPI) networks of cancer associated genes constructed from the corresponding genes by analyzing high-throughput gene expression datasets of human cancer patients may be used for better sensitivity and forecast in understanding the key regulating genes of the corresponding disease
PPI network in prostate cancer (PCa) follows hierarchical scale-free topology composed of modules at five levels of hierarchy

Summary

Introduction

Identification of key regulators and regulatory pathways is an important step in the discovery of genes involved in cancer. We propose a method to identify key regulators in prostate cancer (PCa) from a network constructed from gene expression datasets of PCa patients. The network theory-based methods have been used to predict disease genes from networks generated using curated list of genes reported to be associated with the disease and mapping them to the human gene interaction network (HPRD database)[10] In such approach, the studies have been limited to the curated gene list forming the network not completely representing the system and patient-specific information is not considered. The application of network theoretical methods to the protein–protein interaction (PPI) networks of cancer associated genes constructed from the corresponding genes by analyzing high-throughput gene expression datasets of human cancer patients may be used for better sensitivity and forecast in understanding the key regulating genes of the corresponding disease. The method used in this study is new and takes a holistic approach for predicting key disease genes and their pathways within network theoretical framework using datasets of PCa patients

Methods

Results

Discussion

Conclusion