Tale of two Cambridges: Genzyme, De Novo to apply computational discovery system in collaboration deal

CAMBRIDGE, U.K.—Adding another U.S. feather to its cap, De Novo Pharmaceuticals announced a collaborative agreement with Cambridge, Mass.-based biotech Genzyme that will see the U.K. company apply its proprietary molecular modeling system, SkelGen, to identify potential lead compounds against a Genzyme target. Financial details of the deal were not disclosed.

 

Officials from Genzyme were unavailable for comment, but in announcing the deal, Dr. Thomas Jozefiak, senior scientific director of chemical research at Genzyme, said: "This collaboration will enhance Genzyme's ability to discover cutting-edge therapeutics using a structure-based design approach."

 

The deal is just the latest success of De Novo's decision in December 2003 to refocus the company's efforts on platform commercialization rather than drug discovery, and according to Bill Harris, De Novo's director of drug design, is the direct result of the company's active business development drive in the U.S. Says Harris: "This was a conscious decision to focus on revenue generation through exploitation of our core technology."

 

SkelGen is an automated de novo design program that constructs ligands from a library of 1,700 chemical fragments, which De Novo developed in an earlier collaboration with Roche. According to Dr. Henriëtte Willems, group leader of applied design at De Novo, the program is unlike other modeling systems in that it docks random fragments into the active site of target molecules and using defined connection rules, recombines the fragments to achieve the "best" fit.

 

"This [starting] fragment can be replaced or deleted while the program optimizes the growing ligand by adding fragments to or removing fragments from it, or by rotating, translating or changing the ligand's conformation," she explains. "Another key feature of SkelGen is its ability to take pharmacophoric restraints as input, so that the user can define donor, acceptor and hydrophobic requirements that must be satisfied to within a given tolerance before any structure is generated."

 

In this way, the program should only generate compounds that organic and medicinal chemists can actually synthesize. Willems and colleagues recently showed the efficacy of SkelGen by designing and testing novel ligands that bound tightly to estrogen receptors. They published these findings in a recent Journal of Chemical Information and Modeling.

 

The deal is also reflective of a trend in the biotech and pharmaceutical industries to leverage computational capabilities to generate better drugs less expensively. A recent report by Frost & Sullivan suggests the computational biology market will grow at a compound annual growth rate of 43.5 percent from its current $60 million to $751.8 million by 2011.

 

In announcing the report, drug discovery technology analyst Raghavendra Chitta said: "The rewards of a drug discovery program with a tightly integrated in-silico simulation system are astounding, with the ability to prioritize, validate and eliminate targets at a very early stage in drug discovery. The elimination of false leads at an early stage rather than at the clinical trial stage can offer savings amounting to nearly $200 to $300 million."