Bioinformatics startup Compendia Bioscience added four undisclosed large pharmaceutical customers to its client list in 2007, company officials said this week, which helped revenue for the year grow more than 300 percent.
The company, which spun out of the University of Michigan two years ago [BioInform 03-17-06], markets Oncomine, a database of cancer gene-expression profiles curated from the scientific literature. Oncomine currently contains more than 24,000 transcriptome profiles from around 40 different tissues. The company said that it has derived more than 1,700 gene-expression signatures from the information in the database to date.
John Freshley, Compendia’s chief business officer, told BioInform that 12 of the top 20 pharma companies currently subscribe to Oncomine — up from eight pharmaceutical firms one year ago. AstraZeneca and Merck are the only two pharmas that have allowed the company to name them as customers, Freshley said.
Freshley said that the strong adoption trend in pharma has actually been a bit of a surprise for the firm. “We expected to get them eventually, but I did not expect them to be the early adopters,” he said.
One reason he thinks pharma is turning to the company’s database is “because they understand the problem that we solve, and they’ve been feeling the pain,” he said. “For example, one of the companies we’ve been talking with had assigned close to eight [full-time employees] to try to duplicate what Oncomine was doing, and they gave up after a year after not having made much progress.”
Compendia currently has 14 full-time staffers at its headquarters in Ann Arbor, and partners with the University of Michigan and Harvard University for database curation services. The company also relies on a curation team of 11 people in Bangalore, India.
“One of the companies we’ve been talking with had assigned close to eight [full-time employees] to try to duplicate what Oncomine was doing, and they gave up after a year after not having made much progress.”
Freshley said that the company expects to see the same level of revenue growth in 2008 as it witnessed in 2007 — from new customers as well as expanded agreements with existing customers.
Dan Rhodes, co-founder and CEO of Compendia, said that the company expects to benefit from the trend within pharmaceutical firms to create translational medicine research groups, and noted that the firm launched a new product last year called Oncomine Concepts Edition that is “specifically tailored” to match a drug’s genomic profile “to the underlying biology as well as potential patient populations where that drug would have efficacy.”
Freshley agreed that the use of genomics tools further downstream in the pharma pipeline should be a growth driver for the company. “Early on, in the first year of our business, we were almost exclusively dealing with target-discovery [and] target-validation groups at pharma,” he said. “Much of our growth … has continued to be around target discovery and validation, but also in these translational medicine groups.”
Compendia is also working on a new version of Oncomine that it hopes to launch before the end of the year. The firm received a $2.4 million Small Business Innovation Research grant from the National Cancer Institute last year to help it develop the next generation of the system.
Rhodes said that the primary development goal is to improve the usability of the database. “As the mass of data that we’ve curated continues to grow, and as the experiments become more diverse, it becomes an increasing challenge to find exactly what one is looking for,” he said. “So our major focus early in 2008 is around improving search/browse, as well as the ability to get results out of the system and communicate them to colleagues.”
He said that Compendia also plans to add new data types to Oncomine, which has primarily focused on gene-expression data to date. Rhodes said that the company plans to add array-comparative genomic hybridization data to the database, as well as cell line data.
“Originally, we were largely primary tumor-focused, but what we found is that our pharmaceutical customers are making interesting observations in our primary tumor data, but to act on those observations, they need to model systems to do experimental work,” he said.
Other product-development goals in the year ahead include improving data warehouse to support these additional data types, as well as upgrading the company’s “data pipeline,” an internal software tool that helps its curators capture data from the public domain, annotate it, and normalize it for entry in the database.
Freshley described the pipeline as a “process-management tool” for Compendia’s curators. “A lot of what we do is still really laborious, and to some extent that’s what our competitive advantage is, because we’ve already invested an awful lot of labor in getting the system to where it is and we continue to do so,” he said.