Dataset Information

Neural network cascade optimizes microRNA biomarker selection for nasopharyngeal cancer prognosis.

ABSTRACT: MicroRNAs (miRNAs) have been shown to be promising biomarkers in predicting cancer prognosis. However, inappropriate or poorly optimized processing and modeling of miRNA expression data can negatively affect prediction performance. Here, we propose a holistic solution for miRNA biomarker selection and prediction model building. This work introduces the use of a neural network cascade, a cascaded constitution of small artificial neural network units, for evaluating miRNA expression and patient outcome. A miRNA microarray dataset of nasopharyngeal carcinoma was retrieved from Gene Expression Omnibus to illustrate the methodology. Results indicated a nonlinear relationship between miRNA expression and patient death risk, implying that direct comparison of expression values is inappropriate. However, this method performs transformation of miRNA expression values into a miRNA score, which linearly measures death risk. Spearman correlation was calculated between miRNA scores and survival status for each miRNA. Finally, a nine-miRNA signature was optimized to predict death risk after nasopharyngeal carcinoma by establishing a neural network cascade consisting of 13 artificial neural network units. Area under the ROC was 0.951 for the internal validation set and had a prediction accuracy of 83% for the external validation set. In particular, the established neural network cascade was found to have strong immunity against noise interference that disturbs miRNA expression values. This study provides an efficient and easy-to-use method that aims to maximize clinical application of miRNAs in prognostic risk assessment of patients with cancer.

SUBMITTER: Zhu W

PROVIDER: S-EPMC4195731 | biostudies-literature | 2014

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Neural network cascade optimizes microRNA biomarker selection for nasopharyngeal cancer prognosis.

Zhu Wenliang W Kan Xuan X

PloS one 20141013 10

MicroRNAs (miRNAs) have been shown to be promising biomarkers in predicting cancer prognosis. However, inappropriate or poorly optimized processing and modeling of miRNA expression data can negatively affect prediction performance. Here, we propose a holistic solution for miRNA biomarker selection and prediction model building. This work introduces the use of a neural network cascade, a cascaded constitution of small artificial neural network units, for evaluating miRNA expression and patient ou ...[more]

PMID: 25310846

Similar Datasets

Project description:BackgroundColorectal cancer (CRC) is the third most lethal and malignant type of cancer in the world. Abnormal expression of human microRNA-200a (hsa-miRNA-200a or miR-200a) has previously been characterized as a clinically noticeable biomarker in several cancers, but its role in CRC is still unclear.MethodsThree CRC miRNA expression datasets were integratively analyzed by Least Absolute Shrinkage and Selector Operation (LASSO) and Support Vector Machine-Recursive Feature Elimination (SVM-RFE) algorithms. Nine candidate miRNAs were identified and validated for diagnostic and prognostic capability with the prediction model. The potential roles of the tumor suppressor miR-200a-3p in invasion, migration, and epithelial-mesenchymal transition of CRC cells were elaborated by in vitro studies.ResultsNine miRNAs (miR-492, miR-200a, miR-338, miR-29c, miR-101, miR-148a, miR-92a, miR-424, and miR-210) were identified as potentially useful diagnostic biomarkers in the clinic. The overall accuracy rate of the nine miRNAs in the diagnostic model was 0.94, 0.89, and 0.978 in the testing, validation, and independent validation dataset, respectively. CRC patients in the GSE29622 cohort were separated by the prognostic model into the low-risk score group and the high-risk score group. The area under the receiver operating characteristic curve (AUC) was 0.872 and 0.783 for predicting the 1- to 10-year survival of CRC patients. The performance of the prognostic model was validated by an independent TCGA-Colon Adenocarcinoma (COAD) dataset with AUC values between 0.911 and 0.796 in predicting 1- to 10-year survival. Nomograms comprising risk scores, tumor stage, and TNM staging were generated for predicting 1-, 3-, and 5-year overall survival (OS) in the GSE29622 and TCGA-COAD datasets. Colony formation, invasion, and migration in DLD1 and SW480 cells were suppressed by overexpression of miR-200a-3p. Inhibition of miR-200a-3p function contributed to abnormal colony formation, migration, invasion, and epithelial-mesenchymal transition (EMT). miR-200a-3p binding sites were located within the 3'-untranslated region (3'-UTR) of the Forkhead box protein A1 (FOXA1) mRNA.ConclusionWe developed and validated a diagnostic and prognostic prediction model for CRC. miR-200a-3p was determined to be a potential diagnostic and prognostic biomarker for CRC.

Dataset Information

Neural network cascade optimizes microRNA biomarker selection for nasopharyngeal cancer prognosis.

Publications

Neural network cascade optimizes microRNA biomarker selection for nasopharyngeal cancer prognosis.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets