GlaxoSmithKline’s efforts to fund a 50-person effort with two top universities to verify that the proteins it is picking as targets for drugs are actually involved in the diseases the company hopes to treat are a big example of the Big Pharma dropping its veil of research secrecy.
The new Centre for Therapeutic Target Validation is a collaboration between the London-based drug giant, the Wellcome Trust Sanger Institute, which is one of the top genetics research hotspots in the world, and the European Bioinformatics Institute. The effort will take a “Big Data” approach to pooling genetic data and other information to pick drug targets.
“[Glaxo are] providing people and expertise based analysts with this incredible wealth of expertise,” says Ewan Birney, who will be the interim head of the center. “They've been doing this for a long time and the've had successes and they've had failures and, in some sense, they’ll be bringing those into the public domain. I've been surprised at how engaged GSK has been scientifically and how interesting this problem is to handle.”
Almost all medicines work by blocking or otherwise affecting specific proteins. For instance, statin drugs like Lipitor block an enzyme in the liver that helps the body create cholesterol; aspirin blocks enzymes involved in causing inflammation and clotting blood.
But one big problem for drug researchers is that they often aren’t sure they understand what the proteins they are targeting with experimental medicines do. Problems in the quality of public research, which can be wrong, can make this worse. This project hopes that big data -- looking at information and especially human genetics -- can make the odds of success for new medicines better.
Patrick Vallance, who heads basic research at GSK, says that in the past ten years genetics has gone from being “quite misleading” to “high throughput” and “reliable.”
“Every week there are something like three new rare genetic mutations causing a very rare disease are being identified,” says Vallance. “Every week there is new data coming up on common disease genetic variations. What that does is it creates massive information that is truly about human biology. All of a sudden you get a window into things that are genuinely causal across a range of diseases.”
Of course, fifteen years ago drug company research executives, including those at Glaxo, made similar claims about the sequencing of the human genome and the data produced by a slew of hot gene data companies. That led to a great many failures, some of which took years to read out. For Glaxo, the reminders of its deals with Human Genome Sciences are recent: the lupus drug Benlysta, which worked but has not generated much in the way of sales, and darapladib, a much hyped heart drug that recently failed in a big trial and that also, it turned out, had the unfortunate side effect of making patients smell bad.
But Vallance provides a strong argument that things are different this time. Creating drugs has always been about finding the right proteins to target, he says. Now genetics has become the best way to do that.
Birney insists that genetics provides something that other methods of biology don’t. When you look at a person’s genes, they’ll be for the most part the same throughout life. So when a gene is linked to a disease, it gives a real clue as to how the biology of a disease works. That’s how rare mutations that cause super-high cholesterol led to the cholesterol-lowering statins, and that’s how Birney and Vallance hope that this effort will lead to new drugs.
It used to be that drug companies tried to keep much of their biology research in-house and secret, so that they would start synthesizing promising new compounds before competitors. But this effort puts a lot more of Glaxo’s research out in the open. The bet is that the advantages of having lots of minds working on the problems of drug discovery will outweigh any advantages that secrecy would have provided.
