June 30 marked the beginning of a new phase of opportunity for Human Genome Sciences, as the company regained rights to its cDNA database and technology upon the expiration of a non-exclusive agreement with five partners: GlaxoSmithKline, Merck, Sanofi-Synthelabo, Schering-Plough, and Takeda Chemical Industries.
Not only will HGS be able to use the database and technology internally without obligation to its research partners, but it will also be able to sign new partners. HGS expects its new partnerships to lead directly to the rights to sell drugs, a significant progression from the milestone- and royalty-based partnerships of its past.
However, when the company entered these five partnerships in the mid-1990s, its database was unique: the Human Genome Project was still shooting for a 2005 completion date and Celera Genomics was but a gleam in Craig Venter''s eye. Just how valuable will the company''s data be perceived by potential partners today when there are so many other options available?
According to HGS CEO William Haseltine, the company''s cDNA data is just as unique today as it was in 1995. "What we did, which is so powerful, is to obtain the gene in a form in which it could make its protein," said Haseltine. "That is not the form that you get from the human genome sequence."
Haseltine said HGS has identified 90,000 genesóa count he staunchly defends despite the 30,000 genes estimated by the HGP and Celera in their February papers using gene prediction algorithms. "It''s well known that ab initio computational methods are insufficient to find genes that aren''t previously known to exist for some other reason," said Haseltine. "Their technical method, sequencing the genome, is what we suspectedóa very poor way to find genes. It missed about 70 percent of all the genes, and of the genes that it correctly identified, it missed exons of 70 percent of those."
The 90,000 genes in the HGS freezer and database are based on experimentally derived messenger RNAs. "You can''t do it computationally. If you''re going to make the protein you have to have the messenger RNA," Haseltine said.
And making the protein is what HGS is focusing on these days. With four of its proteins in human clinical trials, the company is now expanding its facilities to accommodate anticipated manufacturing demand for its protein-based drugs. HGS recently opened a three-building manufacturing complex near its Rockville, Md., headquarters, and broke ground on June 20 for the first phase of a new R&D and corporate campus.
But as HGS matures into its intended role as a drug company, it is building more than new facilities. Haseltine said the company''s informatics system must also adapt to handle the influx of new information from R&D, manufacturing, and clinical trials.
Bob Beliveau, formerly director of bioinformatics at HGS, was recently promoted to director of production systems in an effort to modify the company''s existing information systems to accommodate the new sources of data. "Discovering genes isn''t the hard work," said Beliveau. "Developmental requirements and design, manufacturing, FDA regulationóthat''s the hard work."
Beliveau is confident that the company''s current system can adapt relatively painlessly to its new requirements, however. The core of the company''s data was in place by 1996, Beliveau said, and has been expanded to include other types of biological data as further experiments have been run on the company''s samples. Each of the company''s 90,000 sequences is given a "gene identifier" that links it to every piece of information HGS has compiled on itófrom where it''s stored to expression data to patent status and clinical trial performance. HGS also has developed a suite of bioinformatics tools, called CoGenesys, which is part of its integrated system and enables its researchers to run a variety of analyses with the click of a mouse.
Haseltine said the company''s information system was designed to be flexible and expandable. In fact, he said, the "overall information environment" his IT staff developed is used not only for the company''s discovery and development efforts, but also its legal, accounting, human resources, and payroll activities.
"I believe we probably have as good, if not better, internal information systems than anyone in the world," Haseltine said. "I have never seen a better one and I have rarely seen ones that are as good. And when they are as good they''re not as comprehensive."
However, don''t look for HGS to begin peddling its information system, despite what many would see as an opportunity for such a tool. "If it were a big business we think maybe we would look at it, but we think it''s a small business," Haseltine said. "If you look at the total market for those services in the pharmaceutical industry it''s maybe a $250 million to $300 million marketplace. Tiny. One drug can have $1 billion to $10 billion sales. So why be in that business?"
HGS hasn''t ruled out offering its technology to new partners as part of an overall collaboration package, however. Haseltine estimated that the company''s combined suite of tools gives it early discovery and development productivity "20 to 30 times greater, both in terms of dollars, time, and work hours, than that of traditional pharmaceutical companies."
Haseltine said HGS might also consider creating spinoff companies "in a number of different areas" to exploit its technologies without drawing focus from its drug development efforts. "Then, one of the advantages we could bring is know-how of how to transfer an information infrastructure that I think would greatly aid a startup company," he said.