Combining molecular and physiological data of complex disorders

Abstract

The accurate diagnosis of complex disorders is the prerequisite of appropriate and effective treatment. Ideally, the diagnostic process should include the evaluation of molecular and clinic tests alongside medical observations; these clinical observations and laboratory outcomes are usually integrated by an expert physician or one prevails and the other is used for mere confirmation. In clinical bioinformatics, complex datasets are investigated with the aim to improve the clinical management of diseases. The integration of data from disparate data sources is urgently needed and is the prerequisite in the implementation of network medicine which views disease relevant properties as networks and tries to untangle the information content in these networks. Here we developed a graph theoretical framework for combining and untangling the relationships of physiological and molecular data. We then applied the methodology to determine disease related abnormalities of a molecular network derived from serum of patients with schizophrenia and affective disorder. The universality of the concept is demonstrated by an integration of the metabolic data with a standard laboratory test of glucose measurements. The generation of compound networks allows the integrated analysis of disease relevant information and the detection of robust patterns ultimately facilitating the description of complex disorders. This approach could lead to an automatized methodology for improving disease classification and diagnosis. © Springer-Verlag Berlin Heidelberg 2008.