The research of the department is focused on the integrated experimental/bioinformatics research on metabolic networks and enzymatic function. The experimental methods used include metabolome analysis/mass spectrometry, bioinformatics tools for the construction and analysis of metabolic networks, drug target identification, genome annotation, metabolic flux analysis, protein/protein docking and protein structure prediction are developed and used. The enzyme information system BRENDA is curated at the department. It contains the development of experimental methods for the analysis of metabolic profiles, the subsequent application for genotype/phenotype characterization, genome annotation as well as the establishment of an integrated database containing function and property data of biomolecules and organisms based on the BRENDA model. A further aspect is the development and optimization of computer tools for protein drug screening and design.
The metabolic profile of the a number of model organisms, including Corynebacterium glutamicum, Pseudomonas aeruginosa, Pseudomonas putida, Sulfolobus solfataricus, yeast etc. are analysed with respect to their reaction of genetic and environmental modifications, using GC/MS as well as HPLC/MS methods. Combined with parameters available from different databases the information about metabolites, active metabolic pathways and pathway regulation is used for a complete in silico simulation process of the cellular behaviour.
Bioinformatics - systems biology
The tools developed and used are covering the whole pathway from genome annotation and omics results to the prediction of (mainly) microbial cellular reactions to modifications in the genome, the environment like infection of a host, optimization of product formation in a fermenter or treatment with a drug. This includes enzyme function prediction from genome sequences, enzyme reaction classification, metabolic network analysis and simulation, analysis of metabolome data, integration of genome/transriptome/metabolome data, automatic construction of metabolic networks from enzyme annoations data as well as the development of tools for the analysis of mass spectrometry data and the development of tools for protein structure prediction and protein docking.
BRENDA - the enzyme information system
BRENDA is one of the most heavily used databases in the molecular biosciences. It includes more than 2 million enzyme functional and property parameters manually extracted from ca. 70 000 literature references and data from text-mining projects. Data include classification and nomenclature, reaction & specificity, enzyme kinetic data, enzyme structure, isolation & preparation, stability for enzyme and nomenclature, structure and interaction role for enzyme ligands including literature references for all data cited. Today ca. 140 000 visits and more than 5 million hits per month are registered at the BRENDA host. The development of BRENDA is funded by an EU infrastructure grant, allowing it to be free for world-wide access and download.