Listen to this article
This is an experimental feature. Give us your feedback. Thank you for your feedback.
What do you think?
Since the rise of modern medicine, the pharmaceuticals industry has relied on the brainpower of chemists and biologists to discover and develop new drugs.
Their painstaking work has brought about dramatic advances in human health yet the slow pace of progress has prompted a search for new approaches.
As in so many areas, some of the most promising ideas are coming from Silicon Valley. In the era of big data and artificial intelligence, could computer algorithms provide a short-cut to the next generation of medical breakthroughs?
Among the pioneers of computer-based drug discovery is a Californian company called Verseon. It has developed a system for modelling interactions between molecules to accelerate the hunt for compounds that can interfere with disease.
“The bond between molecules is a hardcore physics problem,” says David Kita, Verseon’s co-founder and head of research and development. “To see if a molecule is a good match involves trillions of calculations.”
Until now, finding those matches has been a hit-and-miss process. Drug companies have aimed to increase efficiency with the introduction of “high throughput screening”, which uses robotic laboratory equipment to rapidly test millions of compounds for signs of medicinal potential.
However, the industrialisation of drug discovery has failed to solve big pharma’s productivity problem. On the contrary, the cost of developing drugs has been increasing as easy-to-find compounds have been exhausted. Since 1950, the number of new drugs approved per billion US dollars spent on R&D has fallen by half roughly every nine years, according to research published in the journal Nature.
A report by the Boston-based Tufts Center for the Study of Drug Development in 2014 found that the average cost of getting a new medicine to market had more than doubled in the past decade to $2.6bn.
“The discovery process is completely broken,” says Adityo Prakash, Verseon’s co-founder and chief executive. “High-throughput screening is just a faster kind of trial and error. Occasionally something is discovered through serendipity and then the rest of the industry crowds in to do a variation of the same thing.”
Mr Prakash and his co-founders are aiming to bring greater precision and predictability to the process by drawing on their backgrounds in computing. They had success in the 1990s as founders of the company Pulsent which developed technology to reduce the bandwidth consumed by online videos. The intellectual property, still used in video streaming today, was later acquired by Intel.
“What is the connection between video compression and drug discovery? It’s physics and maths,” says Mr Prakash. Verseon has spent the past 14 years building a computer system capable of exploring hundreds of millions of compounds and modelling their interaction with disease-causing proteins.
“The pharma industry has coalesced around a few million ‘me-too’ compounds that are just a drop in the ocean compared with what’s out there,” says Mr Prakash.
Much remains for Verseon to prove. The company is using its technology to develop its own R&D pipeline. An anti-blood-clotting drug has shown promise in pre-clinical studies and two other programmes, involving cancer and a diabetes-related eye disorder, are still in the discovery phase.
Verseon’s efforts, however, have been given credibility by several high-profile backers including Steven Chu, the Nobel Prize-winning physicist and former US energy secretary. He sits on the company’s scientific advisory board along with Robert Karr and John Leonard, former top scientists at Pfizer and AbbVie, respectively.
One of the company’s biggest investors is Neil Woodford, the prominent UK fund manager. His involvement was among the factors that led Verseon to float on London’s junior Aim market last year, raising almost £66m.
Verseon is not the only Woodford-backed company aiming to add computing power to the search for new drugs. Another is London-based start-up Stratified Medical, which uses artificial intelligence to sift through huge global databases of scientific research in pursuit of hidden patterns. The company’s algorithms have already revealed two new potential drug targets for Alzheimer’s disease, leading to a deal worth up to $800m with an unidentified US drugmaker.
“[Our technology] is being developed by some very good drug developers sitting next to some very talented neuroscientists, themselves sitting next to some very talented mathematicians and computer scientists, all in the same company,” says Ken Mulvany, founder and chairman of Stratified Medical.
Back at Verseon, Mr Kita says pharma has been slower than other industries to embrace this kind of cross-disciplinary working. “Your car was designed by computer. Your aeroplane was designed by computer. The pharma industry has been a holdout against that trend.”
Companies such as Verseon and Stratified Medical are betting that this will begin to change as drugmakers face mounting pressure from investors and from society to increase output from their R&D pipelines at a lower cost.