By Mark Hall
May 7, 2002 | The process of developing new medicines is ludicrously long and expensive — 12 years and $800 million from concept to market, according to a recent report by the Tufts Center for the Study of Drug Development. I know of no other industry in which 12-year development cycles are the norm. Compressing this timeframe has become the battle cry of the drug industry.
The past decade held the promise that high-throughput screening and genomics-based research methods would enable drug discovery companies to develop better, safer drugs more quickly. During the mid-1990s, several companies set five-year goals to halve the drug discovery timeline, triple the number of new chemical entities within development, and decrease drug compound failure downstream. Needless to say, these goals have never been met, at least not in any repeatable fashion. Genomic and proteomic research has placed record numbers of novel targets and leads into the pipeline, but without increasing the capacity of the pipeline to aggregate, analyze, and make critical "go/no-go" decisions based upon the data.
The sheer volume of data flowing out of the new, information-driven research methods, as well as the heterogeneous nature of the data, is a large part of the problem. Even in the most sophisticated labs, much of the informatics effort is still in data integration and interoperability. It will take several years before data management and aggregation efforts give way to the advanced data mining and decision support tools that the industry needs. Nevertheless, several key developments are occurring that should help advance the pace of the next generation of analytic tools.
The IT industry is bearing down on the bio-IT market as one of the few ports of call open for business amidst the rough seas that characterize today's IT marketplace. Many researchers in drug discovery and academia are flabbergasted at the level of interest from IT vendors. Five years ago, their informatics work was conducted largely within the confines of the bioscience community. Nowadays every major IT vendor has launched (or will soon) sizeable "life science" initiatives, and there is a host of smaller vendors who are partially, if not exclusively, focused on supplying solutions for the biomedical informatics market.
Naturally IT companies are determined to win their share of the pot of gold at the end of the rainbow — a $38 billion prize by 2006, according to IDC. But will the flow of new IT products and services into the drug R&D process have a substantive effect upon productivity performance? The greater IT industry has never before been asked to participate as a key partner in an endeavor with such real-life consequences, but they seem to relish the challenge of assisting an effort that will enhance the quality of life.
The IT industry seems to understand the importance of building relationships into the life science community itself, as well as into the cadre of smaller IT vendors who bring a tremendous amount of scientific and drug industry knowledge to the market. The domain expertise and "hooks" that these relationships offer are critical to any vendor's success in the bio-IT market.
Public Resources A high percentage of the informatics resources available today to the life science community come from public sources. Publicly and privately funded research institutions provide software and database resources that are the mainstay of every drug R&D practice worldwide. Most of these institutions have begun to coordinate their efforts, hastening the day when scientists can focus less upon the drudgery of the data and more on what the data have to say.
Moreover, the open source informatics community is as strong and unified as ever. The movement, lead by Ewan Birney of the European Bioinformatics Institute and Lincoln Stein of the Cold Spring Harbor Laboratory, boasts thousands of active developers worldwide and good software that is finding use not just within academic research, but also deep in the drug industry itself. Collaborative efforts across the open source ecosystem are commendable, as witnessed by such community centers as Bioinformatics.Org and the Bioinformatics Open Source Developers Meeting (aka BioHackathon).
Several large IT suppliers have built strong relationships into the open source and academic communities. Because of the free flow of people, products, and ideas between the different communities, academic researchers and open source leaders wield remarkable purchasing influence within the drug industry. Shrewd IT vendors certainly understand this, as well as the wisdom of building brand loyalty early on.
Although we are still several years away from witnessing anything more than modest, incremental improvements in the drug discovery process, the convergent bioscience and IT industries are well suited to reach the next level, even if
|The IT industry is determined to win its share of the pot of gold at the end of the life science rainbow.
not everyone appreciates the timeframe for this evolution. The VC community is still punch-drunk from the last bubbles of the dot-com era, although there are signs of sobriety in relation to the life science opportunity. The open source community and public institutions are in it for the long haul, but will the IT industry stay the course? It is a risky strategy to build solutions focused upon an industry about which they likely know very little. But can they afford not to?
Mark Hall is director of life science research at International Data Corp. and can be reached at firstname.lastname@example.org.