Functional Gene Networks -Advanced Data Management, a Case Study

With the breath-taking progress of public and private genome, transcriptome and proteome initiatives, computer science has to cope with several new tendencies and challenges. Gene and protein sequence collections continue to grow exponentially, and, at the same time, diversity details as well as functional insights are being split among heterogeneous by content and format data sources. Here, the current state of data management remains far behind the needs of a simultaneous activation of sequence related facts for a given biocomputing purpose. To overcome these obstacles, modern achievements in data base technology, real-world-driven data atomization and modelling, the implementation of user-defined functions into the kernel of the database, parallelized algorithms, a smart layer architecture, virtual and materialized views and other appropriate considerations are being incorporated into Kelman's high-end concept of bioinfomatics and functional genome research. These novelties ensure on principle new levels of data consistency and exploitation, and, thus, pave the way to an indepth understanding of gene interplay, involving genes in all their allelic variants, transcripts in different splice-forms and post-synthetic protein processing products. The relevance of distinct molecular versions for health and illness, evolution and economy is subject to the Gene Network concept (http://www.kelman.de). This concept is exemplified here by means of a case study for hereditary disease research: Stringent relations between genes involved in pathologies with hereditary background are discovered by means of a computational pairwise contact mapping approach. The wild-type and mutant products of a hereditary disease gene were found to exhibit differences in their respective set of binding partners, loosing native partners and gaining new partners. Successive computations join all the pairwise contacts into a multiple interlinked network, revealing that the position of disease genes within the network vary in dependence of the mutation considered. By analysing mutation-dependent network alterations, co-players of the disease genes are being proposed as drug target candidates.