Cluster Switches in Gene Expression Data

Following the sequencing of the human genome, the next step is to understand the function of all genes in health and disease. However, experimental study of the functions of all genes in all diseases is impossible and unnecessary, as not all genes are functional in all conditions. However, understanding which genes are functional in each condition and how they are regulated requires a laborious and expensive experimental effort. In this paper we suggest a heuristic framework, CSGI (cluster switching genes identification) for identifying promising genes for thorough analysis. In CSGI we project a cluster defined in one context to its projection in another context, identifying genes that behave differently in different contexts.We provide a case study of immune system clusters showing that our approach identifies clusters representing core conserved biological processes, as well as important genes that switch of clusters.