A Visual Analytics Technique for Exploring Gene Expression in the Developing Mouse Embryo

Abstract

This paper describes a novel visual analytics technique for exploring gene expression in the developing mouse embryo. The majority of existing techniques either visualise a single gene profile or a single tissue profile, whereas the technique presented here combines both - visualising the genes expressed in each tissue in a group of tissues (the components of the developing heart, for example). The technique is presented using data, provided by the Edinburgh Mouse Atlas Project, of gene expression assays conducted on tissues of the developing mouse embryo and a corresponding hierarchical graph of tissues defining the mouse anatomy. By specifying a particular tissue, such as the heart, and a particular stage of development, a Formal Context is computed making use of the hierarchical mouse anatomy so that the resulting Formal Concept Lattice visualises the components of the specified tissue and the genes expressed in each component. An algorithm is presented that defines the computation the Formal Context. Examples of resulting lattices are given to illustrate the technique and show how it can provide useful information to researchers of gene expression and embryo development.