Dataset Information

Relative impact of multi-layered genomic data on gene expression phenotypes in serous ovarian tumors.

ABSTRACT: BACKGROUND: The emerging multi-layers of genomic data have provided unprecedented opportunities for cancer research, especially for the association study between gene expressions and other types of genomic features. No previous approaches, however, provide an adequate statistical framework for or global analysis on the relative impact of different genomic feature layers to gene expression phenotypes. METHODS: We propose an integrative statistical framework based on a sparse regression to model the impact of multi-layered genomic features on gene expression traits. The proposed approach can be regarded as an integrative expression Quantitative Traits Loci approach in which not only the genetic variations of SNPs or copy number variations but also other features in both genomic and epigenomic levels are used to explain the expression of genes. To highlight the validity of the proposed approach, the TCGA ovarian cancer dataset was analysed as a pilot task. RESULTS: The analysis shows that our integrative approach has consistently superior power in predicting gene expression levels compared to that from each single data type-based analysis. Moreover, the proposed method has the advantage of producing a substantially reduced number of spurious associations. We provide an interesting characterization of genes in terms of its genomic association patterns. Important genomic features reported in previous ovarian cancer research are successfully identified as major hubs in the resulting association network between heterogeneous types of genomic features and genes. CONCLUSIONS: In this paper, we model the gene expression phenotypes with respect to multiple different types of genomic data in an integrative framework. Our analysis reveals the global view on the relative contribution of different genomic feature types to gene expression phenotypes in ovarian cancer.

SUBMITTER: Sohn KA

PROVIDER: S-EPMC3906601 | biostudies-literature | 2013

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Relative impact of multi-layered genomic data on gene expression phenotypes in serous ovarian tumors.

Sohn Kyung-Ah KA Kim Dokyoon D Lim Jaehyun J Kim Ju Han JH

BMC systems biology 20131213

<h4>Background</h4>The emerging multi-layers of genomic data have provided unprecedented opportunities for cancer research, especially for the association study between gene expressions and other types of genomic features. No previous approaches, however, provide an adequate statistical framework for or global analysis on the relative impact of different genomic feature layers to gene expression phenotypes.<h4>Methods</h4>We propose an integrative statistical framework based on a sparse regressi ...[more]

PMID: 24521303

Similar Datasets

Project description:Serous ovarian cancer (SeOvCa) is an aggressive disease with differential and often inadequate therapeutic outcome after standard treatment. The Cancer Genome Atlas (TCGA) has provided rich molecular and genetic profiles from hundreds of primary surgical samples. These profiles confirm mutations of TP53 in ~100% of patients and an extraordinarily complex profile of DNA copy number changes with considerable patient-to-patient diversity. This raises the joint challenge of exploiting all new available datasets and reducing their confounding complexity for the purpose of predicting clinical outcomes and identifying disease relevant pathway alterations. We therefore set out to use multi-data type genomic profiles (mRNA, DNA methylation, DNA copy-number alteration and microRNA) available from TCGA to identify prognostic signatures for the prediction of progression-free survival (PFS) and overall survival (OS).We implemented a multivariate Cox Lasso model and median time-to-event prediction algorithm and applied it to two datasets integrated from the four genomic data types. We (1) selected features through cross-validation; (2) generated a prognostic index for patient risk stratification; and (3) directly predicted continuous clinical outcome measures, that is, the time to recurrence and survival time. We used Kaplan-Meier p-values, hazard ratios (HR), and concordance probability estimates (CPE) to assess prediction performance, comparing separate and integrated datasets. Data integration resulted in the best PFS signature (withheld data: p-value = 0.008; HR = 2.83; CPE = 0.72).We provide a prediction tool that inputs genomic profiles of primary surgical samples and generates patient-specific predictions for the time to recurrence and survival, along with outcome risk predictions. Using integrated genomic profiles resulted in information gain for prediction of outcomes. Pathway analysis provided potential insights into functional changes affecting disease progression. The prognostic signatures, if prospectively validated, may be useful for interpreting therapeutic outcomes for clinical trials that aim to improve the therapy for SeOvCa patients.

Project description:Objectives and goals: The causes and molecular pathology of ovarian cancer are essentially unknown. However, it is generally understood that serous ovarian borderline tumors (SBOT) and well differentiated (WD) serous ovarian carcinomas (SC) have a similar tumorigenetic pathway, distinct from moderately (MD) and poorly differentiated (PD) SC. The aim of this study was to identify mRNAs differentially expressed between MD/PD SC, SBOT and superficial scrapings from normal ovaries (SNO),and to correlate these mRNAs with clinical parameters. Results: From the global gene expression analyses, thirty mRNAs differentially expressed between MD/PD SC, SBOT and SNO were selected and validated by RT-qPCR, verifying 21 mRNAs to be significantly differentially expressed (p<0.01). Of these, 13 mRNAs were differentially expressed in MD/PD SC compared with SNO (p<0.01) and were correlated with clinical parameters. VEGFA was significantly upregulated (FC=6.1, p=6.0x10-6), and correlated with progression-free survival (p=0.037). ZNF385B was significantly downregulated (FC=-130.5, p=1.2x10-7), and correlated with overall survival (p=0.03). An increased TPX2 (p=0.03) and FOXM1 (p=0.044) expression correlated with optimal normalization of serum CA125 after treatment. Furthermore, we present a common molecular pathway for MD/PD SC, including VEGFA, FOXM1, TPX2, BIRC5 and TOP2A, all significantly upregulated and interacting with TP53. Conclusions: We have identified 21 mRNAs differentially expressed (p<0.01) between MD/PD SC, SBOT and SNO. Thirteen were differentially expressed in MD/PD SC, including VEGFA and ZNF385B, correlating with survival, and FOXM1 and TPX2, correlating with normalization of serum CA125. By the use of oligonucleotid microarrays (Affymetrix U133 Plus 2.0 Arrays) differences in mRNA expression between eleven MD/PD SC, eight SBOT and four SNO samples were analyzed. Differentially expressed mRNAs were selected for validation by RT-qPCR in additional samples totaling 21 MD/PD SC, 13 SBOT and seven samples of SNO. Each mRNA was run in quadruple. The mRNAs differentially expressed in MD/PD SC compared with SNO were correlated with clinical parameters.

Project description:Ovarian tumors of low malignant potential and low-grade ovarian serous carcinomas are thought to represent different stages on a tumorigenic continuum and to develop along pathways distinct from high-grade ovarian serous carcinoma. We performed gene expression profiling on three normal human ovarian surface epithelia samples, and 10 low-grade and 10 high-grade ovarian serous carcinomas. Analysis of gene expression profiles of these samples has identified 80 genes upregulated and 232 genes downregulated in low-grade ovarian serous carcinomas. PAX2 was found to be one of the most upregulated genes in low-grade ovarian serous carcinoma. The upregulation of PAX2 was validated by real-time quantitative RT-PCR, western blot and immunohistochemical analyses. Real-time RT-PCR showed a statistically significant difference in PAX2 mRNA expression (expressed as fold change in comparison to normal human ovarian surface epithelia) among ovarian tumors of low malignant potential (1837.38, N=8), low-grade (183.12, N=17), and high-grade (3.72, N=23) carcinoma samples (P=0.015). Western blot analysis revealed strong PAX2 expression in ovarian tumors of low malignant potential (67%, N=3) and low-grade carcinoma samples (50%, N=10) but no PAX2 protein expression in high-grade carcinomas (0%, N=10). Using immunohistochemistry, tumors of low malignant potential (59%, N=17) and low-grade carcinoma (63%, N=16) samples expressed significantly stronger nuclear staining than high-grade ovarian carcinoma samples (9.1%, N=263). Furthermore, consistent with earlier immunohistochemical findings, PAX2 expression was expressed in the epithelial cells of fallopian tubes but not in normal ovarian surface epithelial cells. Our findings further support the two-tiered hypothesis that tumors of low malignant potential and low-grade ovarian serous carcinoma are on a continuum and are distinct from high-grade ovarian carcinomas. In addition, the absence of PAX2 expression in normal ovarian epithelia but expression in fallopian tube fimbria and ciliated epithelial inclusions would suggest the potential development of tumors of low malignant potential and of low-grade ovarian serous carcinomas from secondary Müllerian structures.

Dataset Information

Relative impact of multi-layered genomic data on gene expression phenotypes in serous ovarian tumors.

Publications

Relative impact of multi-layered genomic data on gene expression phenotypes in serous ovarian tumors.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure