An Integration Architecture Designed to Deal with the Issues of Biological Scope, Scale and Complexity

This paper discusses a general purpose software architecture, called Addama, which is used to support the rapid integration and analysis of high volumes of complex biological data. It does this by providing: adaptable software which enables interoperable data access; a step-wise and flexible integration strategy, allowing new information to be overlaid on top of existing annotations and context graphs; and through the provision of asynchronous messaging to support rapid integration of new analysis mechanisms. This work is illustrated through the Cancer Genome Atlas (TCGA) study. Addama is being used within a TCGA analysis center to identity new therapeutic intervention approaches by equating clinical outcomes with underlying genomic effects across heterogeneous data from approximately 20,000 patient samples. Addama supports projects like the TCGA through accepting that biological understanding continually changes, and that the rapid integration of new information and analyses is an essential requirement when supporting research.