Predicting the prognosis of breast cancer by integrating clinical and microarray data with Bayesian networks

Abstract

We evaluated three methods for integrating clinical and microarray data: decision integration, partial integration and full integration and used them to classify publicly available data on breast cancer patients into a poor and a good prognosis group. The partial integration method is most promising and has an independent test set area under the ROC curve of 0.845. After choosing an operating point the classification performance is better than frequently used indices.

Prediction of Potential Lead Molecules through Systematic Integration of Multi-omics Datasets - A Mini-Review

“…The network model graphical representation of biological data interrelations and various types of unsupervised dataset integration methods are [44,56]:(i) network-based methodsgraphical representation of interrelations using the network (distance) datasets [45,46],(ii) Bayesian methods -probabilistic graphical representation of interrelations using the probability distribution datasets [47][48][49][50][51], (iii) correlation-based methods -multivariate graphical representation of interrelations using the partial least squares datasets [52,53], (iv) matrix factorization methods -graphical representation of interrelations using the product and rank of the two matrix datasets [54], and (v) kernel-based methods -graphical representation of interrelations using the pattern datasets predicted from kernel matrix [55].…”

Bayesian networks classifiers for gene-expression data

“…For example, Gevaert et al [37] use microarray data and clinical data. They build the Bayesian networks with and without the clinical data.…”

Improved Bayesian Network inference using relaxed gene ordering

“…Candidate graphs with the highest score(s) often serve as the basis for generating new biological hypotheses of the underlying signal transduction mechanisms. BNs have been successfully applied in a variety of biomedical research fields such as cancer biology (Gevaert et al, 2006), biomarker discovery (Diao et al, 2004), system biology (Zou and Conzen, 2004), and genetics genomics (Li et al, 2005).…”

“…Our approach selects subset of context-dependent genes, and infers optimal joint regulatory relationships among genes using a microarray data set describing a specific biological process. The layer constraint makes the NP-complete problem of learning optimal BN structure(s) solvable in a heuristic manner (Gevaert et al, 2006).…”