Start-ups offer chances and pose challenges to young scientists in the bioinformatics environment. Margret Höhe, CEO of the Berlin-based GenProfile AG, gives you an inside view of unique research and working conditions in the start-up environment. GenProfile was founded in 1998 and has already raised about DM 15 million (approx. US$7.5 million) in its first round of financing, making it the largest commercial spin-off from the German Human Genome Project.

GenProfile is based at the Biomedical Research Campus Berlin-Buch, in close vicinity to the Max Delbrück Center for Molecular Medicine (MDC), a national research center, and several university hospitals--one of the fastest growing research areas in the heart of Europe. The start-up's mission is to understand the nature of human genetic variation and its impact on human disease (medical genomics) and treatment (pharmacogenomics). GenProfile focuses on the systematic analysis of DNA sequence variation in biomedically relevant genes in order to identify those variants, or combinations of variants (gene profiles), that determine an individual's genetic risk for disease, predict individually different drug response, and ultimately pave the way to a personalized medicine.

To this purpose, GenProfile utilizes a powerful technology platform, which involves proprietary high-throughput technologies for gene sequence comparison (e.g., multiplex sequence comparison), genotyping (MALDI-TOF minisequencing), unique approaches to data analysis and interpretation, as well as a large network of clinical collaborators. The company's staff, 31 individuals at present, includes molecular geneticists and biologists, mathematicians, scientists with a medical background, engineers, technical assistants, and administrative staff, recruited from well-known German research institutes.

The first working draft of the human genome has now become available. Thus, one of the major future challenges in genome research will be to compare genes and genomes, the analysis of within- and between-species variation. In particular, the comparison of gene sequences in large numbers of patients and controls will be a key step in strategies for disease gene identification. Those few studies that have systematically compared individual gene sequences have shown that genes and the human genome may be much more variable than previously thought. Allelic complexity in genes will be large, and such complexity will make the analysis of genotype-phenotype relationships difficult, particularly in the situation of common, complex diseases. This represents a particular challenge to disease gene analysis: How do we analyze genes and establish genotype-phenotype relationships against a background of high natural genome sequence variability? How do we identify those variants, or combinations of variants (patterns), that are of importance for the phenotype, given that the functionally relevant variants represent only a subset of the naturally occurring sequence variation?

This requires the development of powerful bioinformatic approaches that allow prediction of haplotypes from numerous variants, and the classification of haplotypes into functionally related categories, in order to identify those specific sequence variants, or combinations of variants, associated with the disease phenotype. Bioinformatic approaches to the analysis of genetic variability and complex genotype-phenotype relationships will moreover include gene sequence and database analyses, measures of association of haplotypes/genotypes with phenotype, clustering procedures, neuronal networks, fuzzy and other techniques in pattern recognition, similarity measures for discrete patterns (e.g., gene sequences, structures, functions), logistic regression methods, and a spectrum of other techniques.

The development of "bioinformatic packages" for the analysis of genetic variability and genotype-phenotype relationships will be of great value for any basic and applied comparative genomics research issues. Importantly, genome research and high-throughput data production at an industrial scale will generate a wealth of data, which can serve bioinformatics researchers as a "test bed" for the development and validation of algorithms.

Thus, there is tremendous room for creativity, growth, and development for qualified and highly motivated young scientists--optimally with a background in biology, medicine, or mathematics. Quite a number of additional features appear to be attractive for young people in start-up companies: the much more flat and flexible hierarchies as compared to the more inflexible structures of universities and other research facilities. There is much greater potential for young people to actively shape careers and create long-term perspectives for themselves and their families. Everybody's impact is desired and important for the development and growth of the enterprise; individuals are meant to contribute and make a difference. Needless to say, stock options and more attractive salaries are offered, too.

For a young start-up enterprise as GenProfile, a supportive environment is also tremendously important. The dynamic research environment at GenProfile is complemented by a rapidly improving environment for bioinformatics in Berlin. Jens Reich, one of the first professors in bioinformatics, works in close vicinity at the MDC. The Max Planck Institute for Molecular Genetics, which GenProfile has established very good collaborations with, is in the process of appointing a new department head in this area, planned as a joint appointment with the Free University (FU), and there are also a number of excellent groups in this area in Berlin. Both the FU and Humboldt University are in the process of starting bioinformatics curriculi. Last but not least, research at GenProfile is supported by an excellent international board of senior scientific advisors, namely George M. Church (Harvard Medical School, Boston), Jens Reich (MDC), Axel Ullrich (Max Planck Institute for Biochemistry, Munich), Jürg Ott (Rockefeller University, New York City), and Erich Wichmann (GSF-National Research Center for Environment and Health, Munich).