Files in this item



application/pdfWang_Chunjing.pdf (515kB)
(no description provided)PDF


Title:A systems approach to the different grades and progression of human astrocytoma
Author(s):Wang, Chunjing
Advisor(s):Price, Nathan D.
Department / Program:Chemical & Biomolecular Engr
Discipline:Chemical Engineering
Degree Granting Institution:University of Illinois at Urbana-Champaign
Subject(s):Systems biology
Human astrocytoma
network classifier
Abstract:Astrocytomas are the most common glioma, accounting for about half of all primary brain and spinal cord tumors. Malignancy in these tumors ranges from the least aggressive pilocytic astrocytoma (WHO grade 1) to the most aggressive glioblastoma (WHO grade 4). Molecular biomarkers or signatures—i.e., patterns of gene or protein expression that can reliably distinguish between each grade and provide insight into the underlying molecular events associated with tumor progression—have not yet been well established for astrocytomas. To identify candidate biomarkers and characterize genetic and molecular mechanisms driving glioma development and progression, we performed a meta-analysis of publicly available microarray gene expression datasets, comprising 432 tumor samples from all four grades and 28 non-tumor samples. We first applied a consensus preprocessing method to raw microarray data to reduce bias introduced by different laboratories. Using DIRAC, a network-based classification approach previously developed in our lab, we were able to effectively differentiate tumor grades with an average accuracy of 87%. Additionally, we derived 46 specific transcriptional changes that are associated with astrocytoma progression; of the 46 genes, 27 were consistently upregulated and 19 were downregulated in the progression sequence. Notably, we discovered a histology-independent classifier, a network using erythropoietin to mediate neuroprotection through NF-kB (EPONFKB), consisting of 11 genes and predictive of survival in high grade astrocytoma (HGA) patients. This network is known for its roles in neuronal development and is capable of classifying HGAs into previously unrecognized subtypes. It has proven to be a more significant survival predictor (P = 2.4e-8) than histology-based grading (P = 2.2e-6). With our network signatures associated with each grade and our progression-associated genes, we hope to increase the understanding of molecular mechanisms leading to brain cancer development, maintenance and progression. With the identification of the EPONFKB network as a novel prognostic factor, we hope to move tumor diagnosis and prognosis toward a more quantitative realm.
Issue Date:2011-08-26
Rights Information:Copyright 2011 Chunjing Wang
Date Available in IDEALS:2011-08-26
Date Deposited:2011-08

This item appears in the following Collection(s)

Item Statistics