Sept. 9, 2002 | HE'S BEEN CALLED
|Bio·IT World senior informatics editor Malorye Branca spoke to von Bohlen about his company, the field of bioinformatics, and where he sees the hottest opportunities.
brash, visionary, unpredictable, and focused, but the word that may best fit LION Bioscience AG CEO Friedrich von Bohlen right now is "undaunted." Despite the extraordinary circumstances he's faced in the past year (see "Triumph, Tragedy, and the Aftermath"
), and the current frigid business climate for bioinformatics, von Bohlen is still determined to fulfill his original vision for LION. So far he's been delivering the deals that could make that possible, but plenty of challenges lie ahead. With his new business plan, von Bohlen aims to solve the data integration problem once and for all. Q: What is the root of the bioinformatics industry's current malaise?A:
The frustration comes from there being too many nuts and bolts, too many interfaces that are not
communicating with each other. Because of the venture capital boom in the mid-'90s, during which every line of code was financed with $20 million in venture capital, there are now too many players offering many similar things. It also takes time to translate a cool database like the human genome into something that adds value. It's a technical problem, a science problem, and a business problem.
Information management solutions for life sciences are at least as difficult to build and manage as those for business organizations, and it took SAP at least 10 years to create their business solutions. One of the criticisms for our industry is that here we are, addressing one of the most exciting and challenging aspects of modern drug discovery, and yet, sometimes we give the impression that this is a pretty easy thing to do.
Q: So what's the answer?
A: For us, it's an intelligent blend of products and professional services. There is no one-size-fits-all answer. We start with the most essential part, the database. Other industries don't need to integrate so many diverse databases. But in this field, if you run even the best analysis tools on limited data, you may miss the beef. So we build up from a data-integration platform, adding scientific analysis applications, decision support, and project tracking pools. Some of these tools came out of our work with Bayer, and some were developed in our own drug discovery unit. We will be bringing the best tools together in a modular framework that spans the full range of research and development, so that clients can purchase only the parts they need.
Q: Data integration is a sticky problem. Why do you think LION can conquer it?
A: All the other companies have pieces of the answer, but mine is the only company that has the technology to integrate all data formats and applications. We have SRS to integrate data, and now the Discovery Center technology we acquired with NetGenics helps us to build on that and integrate applications. Creating a hyperlink from application A to B is easy, but that doesn't really get you where you want. It's like buying an airplane engine from one company, wings from another, and bolting them together. When you are done, the thing doesn't necessarily fly. This is an aspect of the problem that very few companies understand. You need to really know science and drug discovery to design integration tools that give people useful results.
Q: How are demands for bioinformatics changing?
A. I think the pharmaceutical industry is going through a transition. They've realized that they have too much from too many vendors. At the same time, for the last year or so the pressure for productivity at pharma has increased substantially. Going forward, they are going to want fewer partners, but these partners will have to have higher capabilities. At the same time, they don't want to give up the applications they've already adopted for specific scientific needs.
Q: What will LION's future role be?
A: There are many players with good tools out there, but our goal is not to be just another company with as many tools as possible. We want to deliver an integration and communication platform that allows researchers to use our technologies optimally and to plug in just the pieces they want. There is a set of best-in-breed tools emerging, some of which will be coming from LION. But science advances everywhere, and there may be important new tools coming from other companies, too. We are seeking to build a system that is flexible enough to bring these tools in, and perhaps we will license and distribute some of these other third-party applications as well.
Q: How important is the $11.7-million Advanced Technology Program (ATP) grant that LION and Paradigm Genetics Inc. were just awarded from the National Institute of Standards and Technology?
A: It's the largest ATP grant ever made for bioinformatics, and it focuses on target assessment, which
|Triumph, Tragedy, and the Aftermath
|The 1997 creation of LION (Laboratories for the Investigation of Nucleotide Sequences) Bioscience AG in Heidelberg, Germany, came at an opportune time.
is critical to pharmaceutical development. The grant will help us look at the quality of the answers you get when you integrate data from multiple sources, whether the data need to be massaged, and how. In the long term, this work could open up a new chapter in the management of diverse and complex biological data, and it will help us facilitate and accelerate some bioinformatics procedures. But you don't run a company on grants, and for the near term it does not mean that much from a business perspective. The grant is a wonderful affirmation, but for financial success we are still aiming for more deals like the one we have had with Bayer, although they may be structured in different ways. Q: How has the deal with Bayer turned out for you?A:
We have delivered Bayer 370 targets and 400 prospective patents around those targets. When one of these targets becomes a drug, we will participate in royalties, but that is several years away. Bayer is pretty happy with our collaboration, so this could advance to something else, such as work finding leads for some of those targets. Since we are allowed to use technology that developed from the deal, we are also capitalizing on that. Q: Why are you selling your drug discovery unit, when so many other companies are employing the opposite strategy?A:
These days there is a tendency for tool companies to turn themselves into product [i.e. compound] companies, because this is what investors seem to like. I think there is a major flaw with that thinking. If you look at the average product attrition rate, in the next 10 years 80 percent of these will fail. That's not a criticism of the companies, it's just a fact about the business. It's naïve to think that doing drug discovery is good and not doing it is bad. Drug discovery companies still need tools that enable them to do better discovery, and that's what we provide. Q: What makes bioinformatics such a high-risk proposition?A:
There is an uncertainty in bioinformatics that stems from the fact that this field is undergoing an evolution. Everyone is certain there must be something big here — the question is how do you crystallize that? I think that LION has the right approach. The Boston Consulting Group study we recently commissioned points to data integration as one of the highest value areas of unmet need in the drug discovery value chain. Our solution will be the one that achieves that.