Email updates

Keep up to date with the latest news and content from BMC Bioinformatics and BioMed Central.

This article is part of the supplement: Proceedings of the Neural Information Processing Systems (NIPS) Workshop on Machine Learning in Computational Biology (MLCB)

Open Access Research

Infinite mixture-of-experts model for sparse survival regression with application to breast cancer

Sudhir Raman1*, Thomas J Fuchs23, Peter J Wild4, Edgar Dahl5, Joachim M Buhmann23 and Volker Roth1

Author Affiliations

1 Department of Computer Science, University of Basel, Bernoullistr. 16, CH-4056 Basel, Switzerland

2 Department of Computer Science, ETH Zurich, Universitaetstrasse 6, CH-8092 Zurich, Switzerland

3 Competence Center for Systems Physiology and Metabolic Diseases, Schafmattstr. 18, CH-8093 Zurich, Switzerland

4 Institute of Pathology, University Hospital Zurich, Schmelzbergstrasse 12, CH-8091 Zurich, Switzerland

5 Institute of Pathology, University Hospital Aachen, Pauwelsstrasse 30, 52074 Aachen, Germany

For all author emails, please log on.

BMC Bioinformatics 2010, 11(Suppl 8):S8  doi:10.1186/1471-2105-11-S8-S8

Published: 26 October 2010



We present an infinite mixture-of-experts model to find an unknown number of sub-groups within a given patient cohort based on survival analysis. The effect of patient features on survival is modeled using the Cox’s proportionality hazards model which yields a non-standard regression component. The model is able to find key explanatory factors (chosen from main effects and higher-order interactions) for each sub-group by enforcing sparsity on the regression coefficients via the Bayesian Group-Lasso.


Simulated examples justify the need of such an elaborate framework for identifying sub-groups along with their key characteristics versus other simpler models. When applied to a breast-cancer dataset consisting of survival times and protein expression levels of patients, it results in identifying two distinct sub-groups with different survival patterns (low-risk and high-risk) along with the respective sets of compound markers.


The unified framework presented here, combining elements of cluster and feature detection for survival analysis, is clearly a powerful tool for analyzing survival patterns within a patient group. The model also demonstrates the feasibility of analyzing complex interactions which can contribute to definition of novel prognostic compound markers.