National Computational Infrastructure

NCI

Facebook Twitter Google+ LinkedIn Pinterest StumbleUpon

Computer-aided drug design

Drug design using the NCI supercomputer

Drug design using the NCI supercomputer

Dr David Wilson from La Trobe University is using Raijin to design new drugs for diseases such as cancer and Alzheimer’s. “We’re looking at kinases – a group of proteins that are involved in the signalling pathways of a whole bunch of different diseases,” explains Dr Wilson. “If we could stop these kinases working, we could target those disease states.” 

Dr Wilson and his team are using molecular modelling programs on Raijin to design small molecules that block the kinases’ active sites. “By understanding the structure of these kinases we can begin to define small molecules that could bind to and block up the active sites of these proteins and stop them from functioning,” he says.

The team is working with experimental biologists to test their computer-designed drugs in the lab. “We model lots of different molecules and then choose the best 10 to synthesise and perform biological assays with,” explains Dr Wilson. “Then, on the basis of those experiments we go back and modify the models to refine the best target molecules. “We’ve been making some good progress and have several lead compounds that look promising.”

The process is complex and compute-intensive, says Dr Wilson. “These molecules and proteins are very dynamic, moving, flexible systems,” he says. “Traditionally scientists thought of molecules binding to active sites like a ‘key in a lock’. But the reason the lock and key at home on your front door works is because the components don’t move – they’re inflexible. If both your key and lock are flexible and can move and change shape, that makes it a whole lot harder. So it’s not quite as simple as making a molecule that will fit to the exact shape of the active site.”

The greater the computing power, the more accurately these molecules can be modelled. “If we had unlimited computing power – and it’s probably what will happen in the future as computers advance – we would repeat our experiments 1,000 times to get really robust statistics of how well a certain molecule binds to the protein and therefore how effective it could be as a drug,” says Dr Wilson.“We can do so much more now even than three or five years ago thanks to the way that computing keeps advancing.”

La Trobe logo

Back to Research

In Collaboration With