Covariate-Related Structure Extraction from Paired Data
In the biological domain, it is more and more common to apply several high-throughput technologies to the same set of samples. We propose a Covariate-Related Structure Extraction approach (CRSE) that explores relationships between different types of high-dimensional molecular data (views) in the context of sample covariate information from the experimental design, for example class membership. Real-world data analysis with an initial pipeline implementation of CRSE shows that the proposed approach successfully captures cross-view structures underlying multiple biologically relevant classification schemes, allowing to predict class labels to unseen examples from either view or across views.
KeywordsPartial Little Square Canonical Correlation Analysis Canonical Variable Covariate Information Data View
We thank Ming Jin, Jin Zhao, Basem Kanawati, Philippe Schmitt-Kopplin, Andreas Albert, J. Barbro Winkler, and Anton R. Schäffner for kindly providing the datasets used in this study.
- Acar, E., Gurdeniz, G., Rasmussen, M., Rago, D., Dragsted, L.O., Bro, R.: Coupled matrix factorization with sparse factors to identify potential biomarkers in metabolomics. In: IEEE 12th International Conference on Data Mining Workshops, pp. 1–8 (2012)Google Scholar
- Barkauskas, D.: FTICRMS: Programs for Analyzing Fourier Transform-Ion Cyclotron Resonance Mass Spectrometry Data. R package version 8 (2012)Google Scholar
- Jamali, M., Ester, M.: A matrix factorization technique with trust propagation for recommendation in social networks. In: Proceedings of the 4th ACM Conference on Recommender Systems, pp. 135–142. ACM (2010)Google Scholar
- Jiang, M., Cui, P., Liu, R., Yang, Q., Wang, F., Zhu, W., Yang, S.: Social contextual recommendation. In: Proceedings of the 21st ACM International Conference on Information and Knowledge Management, pp. 45–54. ACM (2012)Google Scholar
- Pinheiro, J.C., Bates, D.M.: Basic concepts and examples. Mixed-effects Models in S and S-Plus, pp. 3–56 (2000)Google Scholar
- Saunders, C., Gammerman, A., Vovk, V.: Ridge regression learning algorithm in dual variables. In: Proceedings of the 15th International Conference on Machine Learning, pp. 515–521. Morgan Kaufmann (1998)Google Scholar
- Wold, S., Hellberg, S., Lundstedt, T., Sjöström, M.: PLS modeling with latent variables in two or more dimensions. Partial Least Squares Model Building: Theory and Application (1987)Google Scholar