Parkinson's disease may not be a single condition but a collection of biologically distinct disorders, each potentially requiring a different therapeutic approach, according to a new study from researchers at the Vlaams Instituut voor Biotechnologie (VIB) and KU Leuven in Belgium.
Using machine learning to analyse fruit fly models carrying mutations in 24 genes linked to Parkinson's, the team identified two overarching groups and five distinct subgroups of the disease. The work, published in a peer-reviewed journal, challenges the long-held assumption that Parkinson's is a uniform disease with a one-size-fits-all treatment.
Why a single treatment fails many patients
Parkinson's disease, characterised by movement difficulties and progressive neurological decline, affects an estimated 8.5 million people worldwide, according to the World Health Organization. Its prevalence is rising rapidly across Europe and beyond. Yet despite decades of research, effective therapies remain elusive for many patients.
“When clinicians or patients are looking at the disease, they see the clinical symptoms, which unifies people with Parkinson's disease,” said Patrik Verstreken, head of the research group of molecular neurobiology at VIB-KU Leuven. “But when you look under the hood at the molecular level, then you see that they fall into subcategories. And that's important because one drug to target the different molecular dysfunctions in all Parkinson's disease essentially doesn't exist.”
The study's approach was deliberately open-ended. “We came in without any preconceived notions of how a specific mutation would affect our animal model,” said Natalie Kaempf, first author of the study and a researcher at VIB-KU Leuven Center for Brain & Disease Research. “We took animals with mutations in any of those 24 different genes that are causing the disease, and we just monitored their behaviour over periods of time.”
The computer analysis revealed that the different genetic forms of Parkinson's naturally cluster into separate groups. This classification could help scientists search for biomarkers specific to each subgroup and develop drugs aimed at the patients most likely to benefit.
“By having these subcategories, we can now go and look within that group of patients with those particular mutations, search specific biomarkers, and develop drugs tailored to each group,” Verstreken added.
Subgroup-specific treatments show promise
To test the practical implications, the researchers administered a potential treatment to different groups of fruit fly models. The results were striking: a compound that alleviated Parkinson's-like symptoms in one subgroup had no effect on another.
“When we took a first compound that cured subgroup A and tested it in subgroup B, the latter wasn’t rescued,” Verstreken said. “Our study shows that you can make subgroup-specific drugs that have positive effects and are really specific to that subgroup.”
The research is still at an early stage and was conducted in fruit flies, not human patients. However, it points toward a future in which Parkinson's treatments could be matched more closely to the biological cause of a person's disease. The team believes the same machine-learning approach could be applied to other complex diseases driven by multiple genetic or environmental factors.
“The same principle can be applied to other types of diseases. Diseases that are caused by mutations in a variety of different genes or environmental factors could be classified according to this principle,” Verstreken said.
This study adds to a growing body of work using artificial intelligence to refine medical diagnoses. For instance, researchers at ETH Zurich recently demonstrated that AI can predict personality traits from chat history, while a University of Edinburgh study found that cybercriminals find AI tools disappointing. In medicine, AI models now match or beat doctors in complex medical reasoning.
For the millions of Europeans living with Parkinson's, the prospect of personalised treatment offers a glimmer of hope. As the continent's population ages, the need for more effective, targeted therapies will only grow.
