Detection of Genes with Tissue-Specific Patterns Using Akaike’s Information Criterion

One of the important challenges of microarray analysis is the identification of tissue-specific genes whose expression profile is considerably different in particular tissue(s) than in others. Those characteristics facilitate the identification of a large number of possible markers. In general, the problem of identifying tissue-specific expression patterns in multisource data can be viewed as an outlier identification problem. Akaike’s Information Criterion (AIC), introduced almost 30 years ago by H. Akaike, is an information criterion for the identification of an optimal model from a class of competing models [1]. Kitagawa [4] subsequently used AIC to detect outliers and Ueda [5] more recently simplified the procedure. The most significant advantages of those methods are (i) it is possible to reach a relatively objective decision because the procedure does not require the selection of a significance level, (ii) various situations (e.g. single outlier, multiple lowest or highest outliers, two-sidedand grouped cases) can be treated equally. We report the application of a simplified method for the identification of markedly contracting genes from microarray data [2]. The validity of this novel approach is demonstrated by the distribution of the data detected as outliers and by the comparison with the other method.