The proposed research project will apply an information fusion approach to various types of experimental data in order to increase our understanding of the differentiation of human embryonic stem (hES) cells into various specialized cell types. Gene expression profiles from hES cells in different stages of differentiation will be analysed to identify significantly over- and underexpressed genes. The purpose of the analysis is to find genes that might be important in the differentiation process and that are crucial for directing stem cells into specialized cell types. The project will focus on the endoderm development and one issue will be to increase our knowledge about two different types of endoderm development occurring in humans; primitive and definitive endoderm and their derivates. Another issue will be to compare gene expression profiles from cells that lack chromosome 13 with a sub-clone with normal karyotype (Heins et al., 2004). A comparison of gene expression profiles from in vitro derived specialized cells with gene expression profiles from adult cells from the same tissue type will also be conducted. The project will start, however, with an investigation and validation of “housekeeping genes”, which will be used for normalization and calibration of the gene expression levels in the subsequent analyses. The novelty of the project arises from the fact that most previous research in understanding the differentiation of ES cells has been done on animals (Yamada et al., 2002; Stainier, 2002; Asahina et al., 2004), while very little has been done on human ES cells. Since substantial differences both in morphology and in the gene expression pattern are well known between ES cells from these two organisms, it is important to characterize the genetic regulation of the processes responsible for these differences. This project will work with hES cells supplied by Cellartis, a company specialized in hES cell technologies.