By Malorye A. Branca
August 18, 2004 | Reinvention is no rarity in the dynamic life sciences market; in fact, it has become an art form. Now, a unique informatics boutique in New Mexico has hired seasoned executive Andrew DePristo to help Bioreason put the finishing touches on its own makeover and guide the new enterprise to profitability. DePristo joined the company last April as it transformed into a pure-play cheminformatics software provider. Bio-IT World senior informatics editor Malorye A. Branca recently talked to DePristo about why the change was necessary, what will be different, and what is so promising about this particular desert flower and its ClassPharmer software for high-throughput screening analysis.
Q: With all the informatics products that are flooding the pharmaceutical industry, what is the need for ClassPharmer?
A: This is an unusual tool that targets a real pain point in the drug development process. One of the biggest changes in the industry over the past decade has been the rapid adoption and acceleration of high-throughput screening. Everything is coming through in much greater quantity: Chemists are analyzing not one or a few compounds, but hundreds of thousands of molecules in primary screens. But medicinal chemistry is a painstaking process, so you have to speed that process somehow if you want to keep up with the screens.
ClassPharmer automates the data analysis part of screening. Our classification algorithms let scientists and modelers easily organize their data into chemical families, create compound screening lists, generate predictive models, and prioritize compounds. The results are developed from experimental data combined with predictions drawn from chemical structure.
Q: If the software is so good, what went wrong?
A: When Bioreason launched in 1998, it was really hot, and the plan was to participate in drug discovery on a collaborative basis. Companies would send large sets of compounds over, then, for a fee and some royalties on successful compounds, Bioreason would do the analysis that would tell them which compounds to bring forward.
But that business model can’t work. Pharmaceutical companies just don’t hand over the intellectual property they’ve been working on for decades. For every billion-dollar drug, you have hundreds of molecules that are precursors. So the value of these molecules is easily in the tens of millions of dollars. Companies just aren’t comfortable sharing that value with a small data analysis company.
Q: So what’s the new model?
A: About 1.5 years ago the company changed direction, becoming a cheminformatics software company. They stopped trying to get a piece of the IP, and abandoned the collaboration model. Now, we sell analysis tools. The tools are valuable enough that that alone should be sufficient to help companies automate their screening analysis and allow them to make better decisions on what molecules to bring forward. We’re offering a very straightforward return on investment by reducing screening costs. If we can halve the number of molecules you screen, we’re saving you a lot of money.
Best of all, the software is rather inexpensive compared to what it actually costs to do screening. So you are optimizing a very high-priced process, at a fraction of the cost. Of course, if the consulting opportunity arises, we’re always there to help.
Q: Is the new business model working?
A: In less than two years we’ve gotten about 25 customers, and we’re really happy with the mixture. There are a number of big pharmas, including Abbot, AstraZeneca, and Yamanouchi. We’ve also got a lot of biotechs and some nonprofits. And I’m very pleased to say that we are now working with the FDA on developing predictive toxicology tools. So yes, things are turning around and in just the right direction.