CAMBRIDGE, Mass.--When Hoechst and Rhone Poulenc merged last month to form the life sciences company Aventis in Frankfurt, Germany, one of the first steps taken by the new company’s pharmaceutical arm was to spend $40 million to gain total control over the Hoechst-Ariad Genomics Center here.
The center was created in 1997 as a joint venture between Ariad Pharmaceuticals and Hoechst Marion Roussel to pursue functional genomics based on state-of-the-art technologies in molecular and cellular genetics and bioinformatics. As the Aventis Cambridge Genomics Center, the facility will serve as the test-bed for introducing leading edge genomics technologies into the larger Aventis organization.
Rainer Fuchs, who was chief information officer for the Hoechst-Ariad center, is now head of global informatics for Aventis Pharmaceuticals. He will carry out a novel plan to integrate the organization’s bioinformatics and cheminformatics development efforts at the center.
Explained Fuchs, "We are seeing three main data streams--genomics data, high throughput screening data, and combinatorial chemistry data. Drug discovery will become largely an information-driven science. Creating an organizational unit that takes a more integrated approach to dealing with this information deluge will be a competitive advantage for us."
The center’s transition from Ariad to Aventis was an opportunity to reconsider its informatics infrastructure, Fuchs said. Maintaining a traditional structure in which bioinformatics and cheminformatics are separated in different parts of the organization didn’t seem optimal, he told BioInform. "Instead, we decided to create a new unit of informatics that will try to deal with both problem domains simultaneously."
Aventis’s drug discovery model is based on lead generation followed by lead optimization and then product realization, explained Fuchs. His division, Lead Generation Informatics, will deal with data management and analysis problems faced across the entire discovery research process, from chemistry to biology. "Target identification, screening, combinatorial chemistry, and genomics--these are all areas that will have rapidly increasing data streams," Fuchs observed.
Noting that in most pharmaceutical companies biology and chemistry are kept in separate corners, Fuchs called Aventis’s approach innovative.
"Bioinformatics will be the driver for exploring advanced analysis methods for dealing with the problems of making large, heterogeneous databases available to a global corporate network of thousands of scientists worldwide," Fuchs said. Then, a chemistry component would be based on that model.
Fuchs acknowledged that "analyzing sequence data is not the same as analyzing the results of a high-throughput screen," but, he contended that there are two particular areas where there’s "a great opportunity to look at common methodologies."
One is data organization. "How do we integrate various data streams? How do we make information available to our scientists? What are the tools that scientists need to find the particular pieces of information in these ever-growing databases?" At that level, there are similar problems between bioinformatics and cheminformatics, he said.
Another is analysis methods. "Analyzing 100,000 datapoints in a high throughput screen is conceptually quite similar to analyzing 100,000 datapoints in a gene expression experiment. We’re trying to apply advanced statistical methods, mathematical methods, and datamining technologies to those problem domains," Fuchs said.
Asked whether he had trouble persuading Aventis’s directors of the value of an integrating bioinformatics and cheminformatics, Fuchs said, Not really. "There was a clear recognition that the efficient use of information technology will be a differentiating factor between successful and unsuccessful companies. It was just a matter of coming up with a reasonable model for putting it together."