Preclinical ADMETox predictions: Bio-Rad, Aureus, Equbits team up for in silico ADME/Tox

PHILADELPHIA—In early August, Bio-Rad Laboratories announced a collaboration to use specific data sets from knowledge management platform provider Aureus Pharma in conjunction with BioRad's KnowItAll ADME/Tox software, which the companies say will provide improved in silico screening of potential drug candidates.

At the heart of the collaboration are Aureus' curated ADME and hERG knowledge bases, says Gregory Barnik, general manager of the informatics division for Bio-Rad. "We don't have the kind of curated data that Aureus has," Barnik says. "The cornerstone of our business strategy is to partner with companies that have data or tools that will provide added functionality to our products."

For Paris-based Aureus, the collaboration with Bio-Rad comes as the company looks for greater visibility in the United States. Founded in 2000, the company has diligently worked on its mission of gathering and curating published data and the development of custom databases for clients based on a specific theme or therapeutic target.

"We are familiar with the work Bio-Rad has been conducting over the years," says Mary E. Donlan, director of marketing and sales for Aureus. "This is simply a combination of good prediction tools with KnowItAll combined with the good data that we provide."

A third partner in the collaboration is Equbits, a provider of analytic software powered by Support Vector Machine (SVM) technology.

As Barnik sees it, the combination is a perfect triangle. "If researchers don't have their own data, they can use the data from Aureus with Equbits Foresight to create their own model then export the data to KnowItAll to validate the model."

Once validated, researchers can use KnowItAll tools to create literally thousands of structures based on the model and rank order the structures to suggest those most likely to be viable drug candidates based on their ADME/Tox profiles.

To show proof of concept of the power of the using the three products in combination, the companies have created their own models of CYP 2D6 and hERG inhibition.

"The more we can show there is value in this joint collaboration, the better," Barnik says. "We are really excited by these results especially in light of all the recent interest around hERG toxicity and cardiac problems."

The three companies intend to publish the results of these joint research acitivities, though at present they are only available to potential customers as a proof of concept. They will also be available in a future version of KnowItAll strictly for evaluation purposes.

In Barnik's view, a key piece to creating the models was the data provided by Aureus. "The Aureus data is extremely pristine," he says. "The ability to create data in this fashion is a very difficult task and they've done an exceptional job. Models are plagued by data that comes from many different sources and labs. It is an art and a science and it is something they have mastered."