Appearing to swim against the informatics tide, De Novo Pharmaceuticals last week announced that it has abandoned its in-house drug discovery activities to focus solely on the commercialization of its in silico drug discovery platform.
De Novo CEO Burt Wuurman admitted that the company’s new strategy runs counter to the recent trend for tool and technology providers to reinvent themselves as drug discovery firms, but described the move as more of a harbinger of a new era for computational drug discovery. The company’s specialty, computer-based drug design, “is an enormously growing market,” he said. “The number of drug targets has shot up because of the genome project, and it’s likely to be a growing problem for big pharma companies. They’re never going to be able to handle all the targets” using traditional screening methods, he said.
According to Wuurman, it’s “inevitable” that companies will switch to in silico approaches to grapple with the avalanche of targets.
Wuurman joined the company in November from biotech firm Antisoma, where he was director of business development. After a short evaluation of De Novo’s capabilities, Wuurman said he determined that the company’s biggest strength was in its technology platform. “We were working on in-house discovery, but, to be honest, other companies were better at that,” he said. “I decided that it’s better to do what you’re good at, rather than try to be average at what everybody else does.”
Last week, De Novo closed its biology and chemistry laboratories, consolidated its Cambridge, UK, headquarters to a single floor, and laid off more than half of its staff. The company’s 23 remaining employees are focused on “becoming profitable and selling what we have now,” Wuurman said. He expects that the revamped strategy will bring the company to profitability in 2005.
De Novo also reorganized its management structure. Steve Beasley, previously COO, is now chief business officer, and Richard Scott, previously head of cheminformatics, has been made director of technology development.
Wuurman admitted that De Novo will have a bit of an uphill battle as it peddles its technology platform to drug companies burnt by previous generations of over-hyped “rational drug discovery” technology. “In the past, these in silico design platforms were not that sophisticated,” he said. “It’s of course our job to go out there and convince people that we’ve overcome these problems.”
The first generation of rational drug design tools, which appeared in the late 1980s, was indeed a disappointment for many drug discovery firms, agreed an executive from a large pharmaceutical company, but second-generation approaches like De Novo’s “are beginning to make a significant contribution to the pipeline,” he told BioInform.
Just last week, for example, Tripos announced that it successfully designed a lead series for a “high-priority” project at Swedish biotech firm Biovitrum using its suite of software tools. The entire project — from design to synthesis — took less than three months.
“The pendulum is beginning to swing back” in favor of in silico approaches to structure generation, he said. One reason is that the current lead discovery technology of choice, high-throughput screening, has not been able to fill big pharma’s dwindling drug pipeline, he said. In addition, the technology for generating three-dimensional structures has vastly improved over the last decade, so that more lead discovery programs are launched with a high-quality, 3D target structure in hand. Finally, he noted, the software itself has vastly improved, especially in its ability to predict binding behavior.
De Novo’s flagship product is the Skelgen chemical structure generation platform, which uses 21 different physico-chemical parameters to generate chemically tractable molecular structures that bind to a protein’s active site. The approach overcomes a common criticism of earlier computer-based approaches — that “the program would come up with a weird and wonderful molecule that no chemist could synthesize,” Wuurman said, noting that “we’ve solved that problem now.” De Novo’s developers are currently working to improve the technology to account for water molecules on the target binding site and for the flexibility of the target structure.
De Novo has outsourcing partnerships with several chemical synthesis firms so that it can provide a complete solution for its clients, Wuurman said. While Accelrys and other companies offer in silico drug design software, they don’t offer the soup-to-nuts, assisted drug design approach that De Novo does, he said: “We go, for example, to biotech companies who have targets but no chemistry information and we say, ‘Give us your target, and we’ll turn it around and give you white powder back, which you then can take into pre-clinical development.’”
While Do Novo is willing to break from this collaborative model to install its entire technology platform in big pharmaceutical firms that want to use it in-house, “we don’t license our bits and pieces for low value, which is what Accelrys does,” Wuurman said. With biotech companies, De Novo plans to negotiate more flexible terms for collaborations in order to avoid high up-front payments that may scare off smaller firms.
And while the company will “probably have enough cash” to see it to that 2005 profitability date, Wuurman said he’s not ruling out M&A as a future strategy. “We might want to merge with a company that has, for example, strong science and generates targets, or maybe has one or two products in the clinic but doesn’t have an engine to generate molecules, or we could look for another company that has additional platform technology we want to bolt on,” he said.