Detecting signs of this debilitating disease with AI before any bones start to break
Melissa Formosa is an osteoporosis expert at the University of Malta in Msida. She shares her AlphaFold story.
Right now, medicine is too dependent on radiographic imaging techniques for diagnosing osteoporosis. It can be a debilitating disease that develops slowly over several years, weakening bones and rendering them dangerously fragile. Such diagnostic radiographic tools have their own limitations, detecting osteoporosis only once it has already developed. This means that it is already too late to properly control it.
People tend to think of bones as unchanging, but this is a misconception. They are actually highly active organs, made of connective tissue reinforced with calcium and unique bone cells, with most also containing bone marrow in which blood cells are made. Bone tissue is constantly replenished throughout our lives, with these specialised cells absorbing the old tissue and laying down new tissue. These processes work hand in hand to maintain a healthy and strong skeleton.
The complex nature of bones means there are many different mechanisms by which osteoporosis can develop. Injuries from osteoporosis can be very painful and often debilitating, even requiring permanent hospital care in some cases, if someone breaks their back for example.
The disease overwhelmingly affects older women: one in three women over the age of 50 are diagnosed with osteoporosis, whereas one in five men in the same age range will suffer osteoporosis. Research into how and why the disease develops in some people but not in others is often overlooked by scientific research – but it is becoming increasingly clear that osteoporosis has a significant genetic component.
Take the WNT1 gene, which is active in osteoblasts, the cells that creates bone. Mutations in this gene disrupt the process of building bones, meaning that people with this genetic mutation have brittle bones and suffer from early onset osteoporosis. Findings such as these are important in demonstrating that osteoporosis is not – and can no longer be seen as – a disease that solely affects the elderly.
Yet, a fracture is still often the first indication that osteoporosis is present. What we need is to find biomarkers – a blood test or identified gene or protein we can seek out in those who are predisposed, or at high risk of developing osteoporosis. We need to help people start fighting the disease before it’s even begun.
To help us do this, we’ve been using AlphaFold in an attempt to understand genetic causes more fully. We quickly realised that this could revolutionise treatment if we used it to develop personalised medicine. In this case, a model able to provide tailor-made prevention and treatment strategies for defined groups of individuals.
Someone with the disease could then have their genome sequenced. This genetic analysis would give us a better idea if that person is at risk of a fracture in the near future and, crucially, allow preventative action to be taken. It could also help us understand disease progression, and allow patients more control over deciding what intervention is best for them. Ultimately, the aim is to manage osteoporosis from the earliest possible stage, and prevent progression, fractures and the pain and debilitation they bring.
When we enter the amino acid sequence into AlphaFold software, it creates a 3D image of the protein structure and allows us to compare the protein structures encoded by both normal and defective genes. With AlphaFold, we can visualise the impact of specific genetic mutations, some of which may only cause subtle structural changes. Others induce significant deformations to the protein, reducing its ability to function properly, contributing to disease.
Ultimately, we’re aiming to develop simple blood tests for young adults to help predict disease, and to find new genes and proteins associated with the disease so we can develop better drugs to treat it. Early detection and the introduction of personalised medicine could mean that osteoporosis can be managed much more effectively. Millions of lives could be vastly improved.
This kind of AI is becoming central to our work and will be critical to future researchers in this field. We finally have a chance to get one step ahead of this debilitating disease – that’s priceless.