Detecting Signs of Osteoporosis 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.

Currently, medicine relies heavily on radiographic imaging techniques for diagnosing osteoporosis. This condition can be debilitating, progressively weakening bones and making them dangerously fragile over several years. Traditional diagnostic tools have significant limitations, as they identify osteoporosis only after it has developed, which means that effective management is already too late.

People often perceive bones as unchanging, but this is a misconception. Bone tissue is an active organ composed of connective tissue reinforced with calcium and specialized bone cells, many of which contain bone marrow necessary for blood cell production. Throughout our lives, bone tissue is continuously replenished, with specialized cells resorbing old tissue and generating new tissue, maintaining a healthy and strong skeleton.

The complexity of bone biology means there are various mechanisms by which osteoporosis can develop. Injuries related to osteoporosis can be extremely painful and often require long-term hospital care, particularly in cases like spinal fractures.

The disease primarily affects older women, with one in three women over the age of 50 diagnosed with osteoporosis, while one in five men in the same age group will experience the condition. Research into the factors contributing to osteoporosis is often overlooked, but it is becoming increasingly clear that there is a significant genetic component involved.

For instance, the WNT1 gene, active in osteoblasts—which are the cells responsible for bone formation—can have mutations that disrupt bone-building processes. Individuals with such genetic mutations may exhibit brittle bones and experience early onset osteoporosis. Findings like these highlight that osteoporosis is not just a disease of the elderly.

Nevertheless, a fracture often remains the most common initial indication of osteoporosis. Identifying biomarkers—a blood test or specific genes or proteins—will be essential for targeting individuals predisposed to or at high risk of developing osteoporosis, allowing for proactive measures to combat the disease before it manifests.

To advance this effort, we’ve been utilizing AlphaFold to gain a deeper understanding of the genetic causes of osteoporosis. We realized early on that this technology could revolutionize treatment by enabling the development of personalized medicine tailored for specific groups of individuals.

For instance, if someone is diagnosed with osteoporosis, sequencing their genome could provide insights into their fracture risk in the near future. This information would enable timely preventative actions and foster a greater understanding of disease progression, thus empowering patients to choose the best intervention for themselves. Ultimately, our goal is to manage osteoporosis from its earliest stages, preventing progression, fractures, and the associated pain and debilitation.

When we input the amino acid sequence into AlphaFold software, it generates a 3D image of the protein structure, allowing us to compare protein structures produced by both normal and defective genes. AlphaFold helps us visualize the effects of specific genetic mutations, some of which may cause minimal structural changes, while others result in substantial deformations that impair the protein’s functionality, contributing to the disease.

Our aim is to develop simple blood tests for young adults to assist in predicting the disease and to discover new genes and proteins linked to osteoporosis, enabling the creation of better treatment options. Early detection and the implementation of personalized medicine could significantly enhance the management of osteoporosis, leading to improved quality of life for millions.

This type of AI is becoming integral to our research and will be crucial for future scientists in the field. We now have an opportunity to stay a step ahead of this debilitating disease—an invaluable prospect.