|
Data integration within biology is a rather old idea, and as a concept it now influences strongly how complex biological mechanisms are understood and disease etiology revealed. Phenotype-specific information is available in computer-accessible form from a multitude of sources, from molecular level gene-specific knowledge and all the way to healthcare sector data such as electronic patient records. Molecular bioinformatics and systems biology, where properties of genes or networks of genes are revealed, have not been having a large overlap with the medical research communities where results made on the basis of information in patient records, public registries, and from the epidemiological area, are created.
One network biology example described in the talk will be "Heart Development" which involves hundreds of genes, some of which are implicated in congenital heart disease. However, systems-level analyses of how these genes integrate into functional molecular networks, and insight into how these networks are affected in disease, have been is missing. We have combined detailed phenotype information from 255 mouse mutants with high-confidence experimental interactome data to create overviews of the time- and tissue-specific networks coordinating human heart development. The networks were experimentally validated in human embryonic hearts at different developmental stages. The data show that morphogenesis is coordinated by a small set of functional modules that are extensively re-cycled across developmental stages, but each stage is defined by a unique combination of modules. We observe a striking temporal correlation between organ complexity and the number of discrete functional modules coordinating cardiac morphogenesis. The talk will discuss the perspectives within data integration which includes the medical informatics area - both in terms of data and the methods which are used to integrate them. |
|