Revolutionising Cardiovascular Research

Researchers at QIMR Berghofer will begin developing a comprehensive ‘encyclopedia’ of the heart in a world-first project that, if successful, will revolutionise cardiac research and treatment.

It is one of the most ambitious cardiac research projects ever attempted, and is made possible thanks to an $8 million fellowship from the Snow Medical Research Foundation – a philanthropic initiative to support emerging scientists to conduct game-changing research.

The fellowship has been awarded to Associate Professor James Hudson, the head of the Cardiac Bioengineering Research Group at QIMR Berghofer Medical Research Institute.

The newly established Snow Medical Research Foundation (Snow Medical), is the vision of businessman and philanthropist, Terry Snow. The Snow Fellowship targets emerging global research leaders that show the potential to drive, manage and influence the next eneration of health and medical innovation.

The 8-year Snow Fellowship provides outstanding biomedical researchers the independence to focus on building ambitious multidisciplinary research programs and teams capable of changing the face of healthcare in Australia and globally.

What is machine learning?

Machine learning is an application of artificial intelligence (AI) that provides systems the ability to learn and improve from experience without being explicitly programmed.

Machine learning focuses on the development of computer programs that can access data and use it to learn through identifying patterns.

These patterns can be labeled for targeted identification (supervised learning), unlabeled, in which the system looks for whatever patterns it can find (unsupervised learning), or the system can use a ‘trial and error’ algorithm to identify patterns (reinforcement learning).

The amount of data used to train these systems can be enormous – with some systems needing to be exposed to millions of examples to learn a specific task.

However, once the system learns the task, the payoff can be huge, as the system can quickly make related connections that lead to predictions and solutions much faster than a human brain.

Associate Professor Hudson said the development of new heart drugs had stalled worldwide and innovative new approaches were urgently needed. ‘Cardiovascular disease kills more people in Australia than cancer, but the number of drugs currently being developed is about 7-fold lower,’ Associate Professor Hudson said.

‘At the moment we still only understand a fraction of the biology underpinning how the heart works so it’s very difficult to figure out exactly what’s going wrong. And if you don’t know what’s going wrong, you can’t fix it.

‘This research will help us better understand the heart so we may address problems when they arise.’ To conduct the research, Associate Professor Hudson and his team will make 80,000 miniature heart muscles, known as organoids.

‘We know there are about 8,500 genes that control heart cells, known as cardiomyocytes,’ he said. ‘Using tens of thousands of miniature model hearts, we will knock out those genes one at a time and see what the response is. We will then catalogue how each gene controls the biology and function of heart muscle, creating an ‘encyclopedia’.’If they are successful, it could be the most comprehensive database of the heart ever developed. The team will then use ‘machine learning’ to read that information and build a working biological model of heart muscle in a computer.

This model will then allow Associate Professor Hudson and his team, for the first time, to make predictions in a computer about what’s going wrong and how to fix it, before starting the costly and time-consuming process of laboratory experiments.

‘This is a new approach in biology that’s based on engineering principles. If successful, it will greatly speed up research into cardiovascular disease.

‘Our vision is that in future, we will be able to plug in information that tells us about a patient’s genes and environment – quickly compute what is wrong with their heart, and predict with precision what treatment they should receive.

‘In the shorter term, our aim is to use our system to find new targets for more effective drugs for heart failure,’ Associate Professor Hudson said.

This extraordinary approach has never been achieved for any organ, and the hope is it could also be applied to other organs in the future.