Google Goes Cancer: Improving Outcome Prediction for Cancer Patients by Network-Based Ranking of Marker Genes
Open Access
- 17 May 2012
- journal article
- research article
- Published by Public Library of Science (PLoS) in PLOS Computational Biology
- Vol. 8 (5) , e1002511
Abstract
Predicting the clinical outcome of cancer patients based on the expression of marker genes in their tumors has received increasing interest in the past decade. Accurate predictors of outcome and response to therapy could be used to personalize and thereby improve therapy. However, state of the art methods used so far often found marker genes with limited prediction accuracy, limited reproducibility, and unclear biological relevance. To address this problem, we developed a novel computational approach to identify genes prognostic for outcome that couples gene expression measurements from primary tumor samples with a network of known relationships between the genes. Our approach ranks genes according to their prognostic relevance using both expression and network information in a manner similar to Google's PageRank. We applied this method to gene expression profiles which we obtained from 30 patients with pancreatic cancer, and identified seven candidate marker genes prognostic for outcome. Compared to genes found with state of the art methods, such as Pearson correlation of gene expression with survival time, we improve the prediction accuracy by up to 7%. Accuracies were assessed using support vector machine classifiers and Monte Carlo cross-validation. We then validated the prognostic value of our seven candidate markers using immunohistochemistry on an independent set of 412 pancreatic cancer samples. Notably, signatures derived from our candidate markers were independently predictive of outcome and superior to established clinical prognostic factors such as grade, tumor size, and nodal status. As the amount of genomic data of individual tumors grows rapidly, our algorithm meets the need for powerful computational approaches that are key to exploit these data for personalized cancer therapies in clinical practice. Why do some people with the same type of cancer die early and some live long? Apart from influences from the environment and personal lifestyle, we believe that differences in the individual tumor genome account for different survival times. Recently, powerful methods have become available to systematically read genomic information of patient samples. The major remaining challenge is how to spot, among the thousands of changes, those few that are relevant for tumor aggressiveness and thereby affecting patient survival. Here, we make use of the fact that genes and proteins in a cell never act alone, but form a network of interactions. Finding the relevant information in big networks of web documents and hyperlinks has been mastered by Google with their PageRank algorithm. Similar to PageRank, we have developed an algorithm that can identify genes that are better indicators for survival than genes found by traditional algorithms. Our method can aid the clinician in deciding if a patient should receive chemotherapy or not. Reliable prediction of survival and response to therapy based on molecular markers bears a great potential to improve and personalize patient therapies in the future.Keywords
This publication has 66 references indexed in Scilit:
- Experimentally Derived Metastasis Gene Expression Profile Predicts Recurrence and Death in Patients With Colon CancerGastroenterology, 2010
- Estimates of cancer incidence and mortality in Europe in 2008European Journal of Cancer, 2010
- A Gene Signature Predictive for Outcome in Advanced Ovarian Cancer Identifies a Survival Factor: Microfibril-Associated Glycoprotein 2Cancer Cell, 2009
- Modelling prognostic factors in advanced pancreatic cancerBritish Journal of Cancer, 2008
- Gene expression–based survival prediction in lung adenocarcinoma: a multi-site, blinded validation studyNature Medicine, 2008
- Network‐based classification of breast cancer metastasisMolecular Systems Biology, 2007
- Validating Stat3 in cancer therapyNature Medicine, 2005
- The STATs of cancer — new molecular targets come of ageNature Reviews Cancer, 2004
- Gene-expression profiles predict survival of patients with lung adenocarcinomaNature Medicine, 2002
- Gene expression profiling predicts clinical outcome of breast cancerNature, 2002