We design computer algorithms to learn how natural patterns emerge from their underlying component parts.
The natural world is filled with awe-inspiring patterns, think of spirals on an ammonite shell, bird flight formations and the tree-like branching of blood vessels. All of these natural patterns emerge from more fundamental processes in nature. The fundamental components we study in our research include proteins, sections of DNA, lipids and even whole organelles.
First, using the principles of mathematics and physics, the rules of engagement between components are formalized. Then, computer simulations are performed to observe just how patterns emerge, maintain their shape and perhaps even dissolve away again.
There are many examples within clinical research which demonstrate the importance of understanding patterning, for example, monitoring the shape of a growing tumour over time or looking at the different shapes of normal and disease-associated chromosomes during cell division.
Increasingly, we are using the new wave of computational techniques associated with machine learning and artificial intelligence to actually predict the development of patterns. Such work will be key to our understanding of how patterns associated with disease are regulated, with significant clinical implications. For example, we may be able to predict the future size and shape of a patient’s tumour in order to determine more effective courses of treatment.