| Abstract |
Modern science and engineering are becoming increasingly interdisciplinary and in many cases different sets of multivariate data (multiblock data) are collected by very different and highly advanced measurement techniques. This situation has raised the need for data analytical methods that can help to extract information from the large volumes of data, to reveal covariant patterns between the different data blocks and to establish predictive models using the different sources of data. A multiblock situation appears whenever different types of data are collected from corresponding samples. Examples are process technology, where measurements are performed at different stages of processes with the same or different measurement techniques, functional genomics where genetic and phenotypic methods are used to understand the causal chain from genotype to phenotype, and biospectroscopy where highly interpretative spectroscopic methods are applied in order to elucidate biochemical mechanisms. In all theses situations a large number of samples (N) is measured with different techniques and data blocks Xb with (b=1,2,…,B) are obtained for every measurement principle b. The total set of measurements can be described by the supermatrix X=[X1|X2|…|XB] where the single blocks have a row to row correspondence, i.e. the same row is referring to the same sample.
In this chapter we discuss different multiblock approaches for the treatment of multiblock data. As examples we consider data from functional genomics and biospectroscopy in general. The data are obtained from designed experiments. The different multiblock methods are discussed with respect to their capability to establish relationship between the interpretative signals in the different data sets and with respect to their ability to extract information with respect to the design. |
| Referanse |
Kohler, A., Hanafi, M., Bertrand, D., Janbu, A.O., Møretrø, T., Naterstad, K., Qannari, E.M., Martens, H. 2008. Interpreting several types of measurements in bioscience. In: Modern concepts in biomedical vibrational spectroscopy, Lasch, P. ISUP, Inc. dba Blackwell Publishing England pp 333-356 ISBN 978 0 470 22945 3 |
Achim Kohler
Trond Møretrø
Kristine Naterstad
Harald Martens
