No cure for diabetes. No cure for Aids. No cure for Alzheimers or the common cold. No new drugs except those for weight or sexual performance.



That is the likely outcome of 40 years of rising costs and longer times to get new drugs certified.



Howard Asher, director of global life sciences for Sun Microsystems, said that when he started in the industry in 1969, it took US$6 million and eight years from discovery to bring a new drug to market.



“Today it takes 17 years and US$800 million. If this continues, once it reaches 20 years and US$1 billion we are out of the business of drug innovation and drug discovery, because 20 years eliminates all worldwide patent protection and eliminates the return on investment that only blockbuster drugs would even come close to offering.


“It means in five to seven years from today, the only survivors would be generic drug companies and the only products they would have … would be those drugs we currently have on the market.”



Asher says the solution is silicon. By finding new ways to harness huge amounts of computing power, the costs and time of drug testing can be reduced.



Mapping the human genome has given researchers a massive amount of data which can be used to study reactions to disease and therapeutic interventions at a molecular level.



“That detail is incredibly promising but it is computationally huge.”



He said the DNA of the animals commonly used for testing had also been mapped, meaning it should be possible to simulate tests “in silico”.



Information on the way existing drugs worked on individual patients could also be collected and compared against genetic data.



“We believe in 10 years we can eliminate the need for all animal studies, we can eliminate phase one and phase two clinical studies, so computationally we can model a drug or therapeutic agent in the computer against the genomic data.”



Asher said new tools and the development of grid computing utilities, which use broadband internet to tie together multiple computers so their CPUs can become part of a virtual supercomputer, could bring the price of getting new drugs released to under $100 million and 10 years.



Researchers need tools to sift the huge amounts of unstructured data – text information in any language or jargon – which may relate to their project. With a terabyte of textual information being published every day in medical, clinical and scientific journals, existing databases and classification systems can’t cope.



Asher said the tools developed by Hamilton firm Reel2 appeared to have solved the problem of finding unstructured information in any language from large databases. “What Reel2 promises is revolutionary for the life sciences and medical communities.”



He said Sun was tackling the problem of how life sciences researchers could amass the computing power they needed without breaking the bank.



In Canada Sun was developing a computing utility.



“Canada has a lot of companies with very large computers but many sit idle on evenings or weekends.



“In life sciences we have people in the post-genomic era who have computational needs which can take days to get from their own systems. It seems insane for them to buy immense systems to manage the peak. A utility will allow them to buy CPU power when they need it.”



The company unlikely to benefit from the explosion in life sciences computing is Microsoft.



Asher said that was because regulatory agencies demanded uncorrupted data – something hard to guarantee from a Microsoft operating environment.



“The FDA [United States Food and Drug Administration] has told me it is concerned at operating systems which decay with use, that under computational stress have a habit of crashing. When operating systems crash they corrupt, and then the FDA says ‘If this data crashed, prove it didn’t corrupt.’ That’s a tall order.”



Asher said that when a problem emerged with the performance of Excel under a new Pentium processor because of a problem with a floating decimal point, the FDA returned 10 new drug applications to sponsors.



“Only seven came back. The other three lost their statistical significance. They estimate that cost the industry and the public over US$6 billion.”
More here.