| Affiliation: | (1) BG Medicine Inc., 610 Lincoln Street North, Waltham, MA 02451, USA;(2) VTT Technical Research Centre of Finland, Tietotie 2, PO Box 1500, Espoo, 02044 VTT, Finland;(3) TNO Quality of life, PO Box 360, Zeist, 3700 AJ, The Netherlands;(4) Center for Medical Systems Biology, LACDR, Leiden University, Gorleaus Laboratories, Leiden, 2300 RA, The Netherlands;(5) Data Theory Group, Faculty of Social and Behavioral Sciences, Leiden University, PO Box 9555, Leiden, 2300 RB, The Netherlands;(6) Nonlinear Dynamics of Natural Systems, Mathematical Institute, Leiden University, Niels Bohrweg 1, Leiden, 2333 CA, The Netherlands;(7) Department of Statistics and Stanford Linear Accelerator Center, Stanford University, Stanford, CA 94305, USA;(8) Biosystems Data Analysis, Faculty of Sciences, University of Amsterdam, Nieuwe Achtergracht 166, Amsterdam, 1018 WV, The Netherlands |
| Abstract: | A novel clustering approach named Clustering Objects on Subsets of Attributes (COSA) has been proposed (Friedman and Meulman, (2004). Clustering objects on subsets of attributes. J. R. Statist. Soc. B 66, 1–25.) for unsupervised analysis of complex data sets. We demonstrate its usefulness in medical systems biology studies. Examples of metabolomics analyses are described as well as the unsupervised clustering based on the study of disease pathology and intervention effects in rats and humans. In comparison to principal components analysis and hierarchical clustering based on Euclidean distance, COSA shows an enhanced capability to trace partial similarities in groups of objects enabling a new discovery approach in systems biology as well as offering a unique approach to reveal common denominators of complex multi-factorial diseases in animal and human studies. Doris Damian, Matej Orešič, and Elwin Verheij contributed equally to this work. |
