On Information Fusion in the Life-Sciences

Abstract

In the present text, we shall discuss soft computing as concerned with the fusion of qualitative percepts and quantitative measurements. Whereas the former is obtained from observation, we shall assume that the latter is collected through measurements. Both, observation and measurement provide the information from which we derive knowledge in the modern life-sciences. Studying gene and/or protein interactions, biological databases are used to combine, for example, time-series measurements with sequence data and information about the classification of genes. These classifications or ‘annotations’ are the result of other experiments and observations, not necessarily linked to the experimental data investigated. The fusion of such context-dependent information with facts extracted from numerical data is of importance in the area of bioinformatics. The paper is to highlight the problems of information fusion in the life-sciences and outlines a conceptual framework in which to formulate such problems.

The support of the UK Engineering and Physical Sciences Research Council (EPSRC) under grant GR/N21871 (Genetic Systems Network) is acknowledged.