T R Hvidsten | PubMed Health

PubMed2001

[New knowledge derived from measurement of gene expression with the DNA microarray method].

BACKGROUND: The cDNA microarray method offers the first possibility of obtaining a global understanding of biological processes in living organisms, by simultaneous read-outs of tens of thousands of mRNAs. Initial experiments suggest that genes with similar function have similar expression patterns. MATERIAL AND METHODS: Understanding this level of biological complexity will, however, require completely new approaches to data analysis. Computer science methods, such as data mining and knowledge discovery, can synthesize interpretable if-then rules that model the relation between gene expressions and functions and use the rules to classify unknown genes. The huge body of existing biological and medical knowledge makes it necessary to develop methods for extracting knowledge from such repositories. RESULTS: Models of relations between gene expressions and gene functions in a data set from a publicly available source are synthesized semiautomatically and applied to classify unknown genes. Encouraging results have been achieved. The method is applied in the analysis of data from our microarray system which has recently become operational. INTERPRETATION: The principles are of general importance and will be used to evaluate a wide range of complex data sets like decision support in clinical medicine, for situations in which physicians need to handle a large volume of data for each patient.

Computational Biology↗

PubMed2001

Predicting gene function from gene expressions and ontologies.

We introduce a methodology for inducing predictive rule models for functional classification of gene expressions from microarray hybridisation experiments. The basic learning method is the rough set framework for rule induction. The methodology is different from the commonly used unsupervised clustering approaches in that it exploits background knowledge of gene function in a supervised manner. Genes are annotated using Ashburner's Gene Ontology and the functional classes used for learning are mined from these annotations. From the original expression data, we extract a set of biologically meaningful features that are used for learning. A rule model is induced from the data described in terms of these features. Its predictive quality is fine-turned via cross-validation on subsets of the known genes prior to classification of unknown genes. The predictive and descriptive quality of such a rule model is demonstrated on the fibroblast serum response data previously analysed by Iyer et. al. Our analysis shows that the rules are capable of representing the complex relationship between gene expressions and function, and that it is possible to put forward high quality hypotheses about the function of unknown genes.

Algorithms↗