Collaborative Drug Discovery said this week that it has received a $1,896,923 grant from the Bill & Melinda Gates Foundation to create a cheminformatics database and expand its software functionalities for scientists developing therapies for tuberculosis.
The company said that CDD will provide services relating to software and computational infrastructure, and it will work with academic researchers and commercial drug developers who will provide chemical and biological data.
The new TB database will complement another Gates Foundation-supported tuberculosis resource called the TB Database, which was developed by researchers at Stanford University, the Broad Institute, and others. Integration between the two resources is already underway.
While TBDB is primarily a genomics resource that houses genes, targets, and microarray data, the CDD database is focused on drugs, targets, and chemical data, so "the connection point [between the two databases] is the target," Bunin said.
CDD, based in Burlingame, Calif., offers a web-based software platform for researchers to archive and mine their results from small-molecule screening as well as data from cell and animal experiments. Researchers who use the company's tools can either share their datasets or choose to keep them private, CDD CEO Barry Bunin explained to BioInform. The new TB database will be built upon this existing infrastructure.
Bunin said that the award process for the TB database was "highly competitive" and that the firm won the grant over another vendor in the space whom he declined to name.
Gates Foundation officials did not comment by press time, but according to Bunin, CDD landed the grant because while the foundation had seen others propose similar projects, they had "not seen [them] implemented as well as CDD has done."
Bunin said that CDD considers its primary advantage in the area of drug discovery informatics to be its "unique collaborative capabilities," which let researchers choose to keep their data private, selectively share it within consortia, or open it up completely.
Other advantages are "ease of uploading chemical and biological data, proven community building capabilities, and superior data, software, and scientific support, which is as important as the database for success," he said.
An Agnostic Platform
The CDD tuberculosis database is being built with input from the scientific community. The software is open to any researcher in the tuberculosis area for free, "courtesy of the Gates Foundation, and for any other therapeutic areas on a commercial basis," said Bunin.
In gathering this feedback, the company deploys "best software practices and robust automated testing suites" to minimize any defects or bugs upfront, Bunin said. "This costs us more initially to do it right, but pays off in performance and scalability of the application to other needs."
CDD's underlying technology is "therapeutic-area-agnostic," he said. "We are focusing it around tuberculosis now."
"The lowest common denominator capabilities are in the application already for chemical and biological [data] mining," he said. The tuberculosis database will let users look for a compound for a particular target, or will allow users to store their own data from low- or high-throughput screening. Compounds can be searched by chemical structure, potency, selectivity, toxicity values for therapeutic windows, or via Lipinski properties, he said.
CDD has also set up a computational development group that will create more advanced functionalities for scientists developing TB therapeutics. "Obviously it is all hands on deck for TB now," Bunin said. Some computational work will be specific to TB, but many capabilities are therapeutic-area agnostic, such as like ADME/Tox predictions, he said.
[ pagebreak ]
As part of this effort, the firm recently hired Sean Ekins as director of collaborations. Ekins, who most recently served as vice president of computational biology at GeneGo, previously held computational chemistry roles at Pfizer, Lilly Research Laboratories, and Concurrent Pharmaceuticals, and will "provide computational services not in the product," Bunin said.
The first step in developing the new database will involve reaching out to TB researchers in academia and industry and "cataloging what specific things they want and need," Bunin said.
He added that the firm intends to work with several of its existing pharma customers on the project. "We have worked with them commercially for other therapeutic areas but the Gates [Foundation] is subsidizing it for TB," he said.
Another level of database functionality undergoing fine-tuning is lending the CDD platform "more fine-grained access control," he said.
"Every object in the databases is tagged in terms of whether it is meant for private [use], for one group, shared for multiple groups, or open. This can be done at the level of an individual molecule or individual bio-data point," he said.
CDD's business model is as an application service and software-as-a-service provider. After it was founded in 2004, the company worked mainly with academics, but now the firm is working " increasingly with foundations and biopharmaceutical companies as they start to do things more collaboratively," Bunin said.
In the "past few months," half of the company's deals have been with industry and half are with academic groups. "It's been a real shift in the tides," he said.
Bunin said that CDD, which is privately held, has increased its user base more than seven-fold over the last year but did not offer further details regarding revenues or customers. "We are doing very well; even in the down economy we are aggressively hiring."
One Database Meets Another
This new CDD TB database will join another Gates Foundation-funded resource called the TB Database, which has more of a genomics focus than the CDD project.
TBDB offers access to tools and resources at the Stanford Microarray Database and the Broad Institute, including microarray, RT-PCR, and genomic data on Mycobacterium tuberculosis. It offers annotation data from resources such as BioHealthBase, a biodefense database on five pathogenic organisms from the National Institute of Allergy and Infectious Diseases; the Institut Pasteur's resource Tuberculist; SRI International's BioCyc pathway and genome database; and additional resources from the Max Planck Institute for Infection Biology and the M. tuberculosis Structural Genomics Consortium.
The CDD database "is, from my perspective, compound-based and is focused specifically on drug discovery," Gary Schoolnik, principal investigator of the TBDB project, told BioInform in an e-mail.
Schoolnik, chief of the Division of Geographic Medicine and Infectious Diseases, Microbiology, and Immunology at Stanford University School of Medicine, said that the TBDB team is "working collaboratively with [the CDD] group and are developing links between our site, TBDB.org, and their site."
TBDB "focuses on a broader agenda of activities" than the CDD effort that are aimed at deepening the biological understanding of M. tuberculosis in the hopes of propelling drug development, the development of new vaccines and biomarkers, Schoolnik said.
[ pagebreak ]
"Our approach is fundamentally driven by genomics, whereas the CDD approach is focused on compounds, their structures and activities. The two sites we both believe are very synergistic with each other," he said.
"We've done some prototypes and this is pushing the envelope a bit in terms of data from really different worlds," CDD's Bunin said in describing the link under construction between the new TB database and TBDB.
Compounding the Information
The new CDD database stands to give academics "access to an important resource [that is already] enjoyed by drug companies," Carl Nathan, chairman of the department of microbiology and immunology at Weill Medical College of Cornell University, told BioInform in an e-mail.
While TBDB is geared toward accumulating and searching genomic information, "CDD is geared toward the results of testing large numbers of chemical compounds for their ability to kill M. tuberculosis under various conditions — that is, early-stage TB drug discovery," Nathan said.
"In the pharmaceutical industry, companies pay large fees to access software programs for keeping track of and analyzing the results of chemical screens," he said, but in the not-for-profit world, "it is rarely possible to access advanced cheminformatics."
CDD, he noted, "is devoted to providing that capacity to not-for-profits, but further, makes it possible for them to share results with each other."
Nathan is a subscriber to the CDD database. At first he purchased subscriptions, one per lab member. "Now that CDD has a Gates grant for outreach to labs like mine, I don't have to pay a subscription fee," he said. Nathan is also a collaborator on the TB database since his team makes suggestions to CDD. "When these work out well, CDD is incorporating them into the service available for other users. There is no financial relationship in either direction," he said.
In a 2007 paper in Nature Medicine "Aligning pharmaceutical innovation with medical need," Nathan outlined more than 20 ways to accelerate the development and distribution of vaccines and drugs for populations in countries that cannot afford them.
One challenge in emerging public-private partnerships, Nathan noted in the paper, is that they must take a "layer-cake approach" to drug development in which academics carry out early research but lack access to compound collections, screening facilities, or pharmacologists, and must hand off "what are likely to be suboptimal lead compounds."
Nathan told BioInform that the CDD project is a step toward improving this scenario, but doesn't completely address the challenge.
"While CDD gives academics access to an important resource enjoyed by drug companies, academics [still] lack many other resources that are essential to drug development — including large-scale funding and diverse kinds of specialized expertise," he said.
"It is exceedingly unlikely that the essential gateway provided by CDD will by itself fill the pipeline for TB drug development. Yet this key step undoubtedly will help put pressure on society to find a way to take subsequent steps," he said.