Computational Identification of Gene Networks as a Biomarker of Neuroblastoma Risk.

Lidan Sun,Hong-Gang Wang,Libo Jiang,Claudia Gragnoli,Christa N Grant,Rongling Wu,Zhenqiu Liu

doi:10.3390/cancers12082086

Abstract

Neuroblastoma is a common cancer in children, affected by a number of genes that interact with each other through intricate but coordinated networks. Traditional approaches can only reconstruct a single regulatory network that is topologically not informative enough to explain the complexity of neuroblastoma risk. We implemented and modified an advanced model for recovering informative, omnidirectional, dynamic, and personalized networks (idopNetworks) from static gene expression data for neuroblastoma risk. We analyzed 3439 immune genes of neuroblastoma for 217 high-risk patients and 30 low-risk patients by which to reconstruct large patient-specific idopNetworks. By converting these networks into risk-specific representations, we found that the shift in patients from a low to high risk or from a high to low risk might be due to the reciprocal change of hub regulators. By altering the directions of regulation exerted by these hubs, it may be possible to reduce a high risk to a low risk. Results from a holistic, systems-oriented paradigm through idopNetworks can potentially enable oncologists to experimentally identify the biomarkers of neuroblastoma and other cancers.

Highlights

Neuroblastoma, one of the most common cancers in children, is highly complex in terms of its genetic, physiological and clinical heterogeneity
Single gene analysis has proven its power for risk stratification, we we argue that neuroblastoma risk includes multiple genes that interact with each otherintricate through argue that neuroblastoma risk includes multiple genes that interact with each other through intricate but coordinated interactionTraditional networks. approaches
We have developed a computational model to reconstruct and implement fully informative gene networks as the biomarkers of neuroblastoma risk

Summary

Introduction

Neuroblastoma, one of the most common cancers in children, is highly complex in terms of its genetic, physiological and clinical heterogeneity. It is this complexity that makes it extremely challenging to diagnose when and how neuroblastoma develops and further design the precise intervention [1,2]. Several clinical parameters, such as age at diagnosis, tumor stage, genomic amplification of MYCN oncogene, and ploidy, are widely used as the markers of neuroblastoma risk [3,4,5,6]. As a widely used approach for general cancer research, such reductionist thinking can simplify the identification of key major genes for neuroblastoma risk [7]. By regarding neuroblastoma as a network disease, we can develop and apply a holistic, systems-oriented approach to better reveal and interpret its genomic causes [11,12]

Methods

Results

Discussion

Conclusion