Crunching the numbers on epilepsy

With the help of an IBM supercomputer, based at the Victorian Life Sciences Computation Initiative (VLSCI), researchers are on the path to developing personalised medicine for epilepsy sufferers.

For the first time, researchers have identified that depression may be a predictor for how well epilepsy sufferers respond to medication.

Dr Slave Petrovski of the University of Melbourne and the Royal Melbourne Hospital says this is the first time that mental health status has been proposed as a useful predictor for anti-epileptic treatment outcome.

“Currently there is no standardised approach for predicting treatment outcomes,” he says, “so it can at times feel like a trial and error process for all involved.

“But what our studies have found is that the patient’s pre-treatment mental health status correlates with their treatment outcome. That is, patients who have a better neuropsychiatric score, which measures levels of depression and anxiety, are more likely to experience initial freedom from seizure when they commence their anti-epileptic medication.”

Dr Petrovski’s research involved investigating 245 newly treated sufferers of epilepsy, combining mood questionnaires, MRI scans and the genome profile of the patient.

“By combining these different sources of information as one integrated model, we have improved the predictive clinical value over what any individual piece of information could have provided alone,” he says.

“These findings suggest that the effective treatment of mood disorders prior to or concurrent with beginning anti-epileptic therapy, has the potential to improve treatment outcome, and reduce seizure recurrence in patients with underlying mood disorders.

“Being able to provide a prediction model based on a combined approach of assessing clinical factors such as mental health status, genome analysis, and brain imaging could revolutionise epilepsy treatments,” he says.

In the next stage of the research, Dr Petrovski and colleagues within the Epilepsy Pharmacogenomics team and experts at the VLSCI are using the computational capacity of an IBM Blue Gene supercomputer based at the VLSCI to conduct a wider genome analysis of data from a larger group of 1500 newly diagnosed epilepsy patients.

“To be able to exploit the compu- tational power of a Blue Gene supercomputer enables researchers to conduct more complex genetic analyses and in particular is helping us to develop a more precise predictor of effective treatments for epilepsy sufferers,” Dr Petrovski says.

“Ultimately our goal is to provide personalised prescribing to patients who suffer epilepsy, so they have more assurance and confidence in their treatment, and with the help of the supercomputer, we are on the way.”

The University of Melbourne and IBM have also announced the installation in 2012 of a Blue Gene/Q supercomputer, which is ranked the greenest in the world.

Expected to be operational by June, the IBM Blue Gene/Q will provide the equivalent computing power of more than 20,000 desktop computers – making it one of the fastest supercomputers in Australia and the fastest supercomputer dedicated to life sciences research in the southern hemisphere.

https://www.vlsci.org.au/