Researchers have used machine learning to identify five distinct molecular subtypes of Parkinson’s disease, challenging the view of it as a single disorder and explaining why treatments fail to work uniformly across patients.
A new study has discovered that Parkinson’s disease can be divided into distinct subtypes, helping researchers to understand why treatments often fail to work uniformly across patients and potentially leading to new personalised therapies.
Researchers from VIB and KU Leuven used machine learning techniques to identify two main groups and five subgroups of the condition, challenging the idea that Parkinson’s is a single disease.
Rethinking a complex disease
Parkinson’s disease affects millions worldwide, defined by symptoms such as movement difficulties and progressive neurological decline. However, scientists have recognised that the condition can stem from mutations in many different genes, resulting in a wide range of underlying biological mechanisms.
This complexity has made it difficult to develop effective treatments. Therapies that target one specific pathway may not work for patients whose disease is driven by a different mechanism.
Parkinson’s disease affects millions worldwide, defined by symptoms such as movement difficulties and progressive neurological decline
The new research suggests that these genetically diverse forms of Parkinson’s can be grouped into distinct molecular subtypes. This changes how the disease is seen, as a collection of related conditions rather than a single disorder.
“When clinicians or patients are looking at the disease, they see the clinical symptoms, which unifies people with Parkinson’s disease,” says Professor Patrik Verstreken at the VIB-KU Leuven Center for Neuroscience. “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.”
Letting the data lead
To uncover these subtypes, the research team adopted an unbiased approach. Instead of starting with assumptions, they studied fruit fly models carrying mutations in Parkinson’s-related genes and monitored their behaviour over time.
Using computational analysis and machine learning, they identified natural patterns in the data.
“We came in without any preconceived notions of how a specific mutation would affect our animal model,” said Dr Natalie Kaempf, first author of the study. ”We took animals with mutations in any of those 24 different genes that are causing the disease and we just monitored their behavior over periods of time.”
This method allowed the team to detect previously hidden structures within the disease and discover how different genetic forms cluster into distinct subtypes.
Towards personalised treatments
The findings could help to inform future treatment strategies. By identifying specific subgroups, researchers may be able to develop therapies tailored to particular molecular profiles.
“We now know that there are different kinds of Parkinson’s disease,” said Verstreken. “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.”
By identifying specific subgroups, researchers may be able to develop therapies tailored to particular molecular profiles
In experimental models, the team tested compounds on different subgroups and found that responses varied. Some treatments were effective for one subgroup but not for another.
“When we took a first compound that cured subgroup A and tested it in subgroup B, the latter wasn’t rescued. Our study shows that you can make subgroup-specific drugs that have positive effects and are really specific to that subgroup,” Verstreken concluded.
Broader implications
The researchers believe this approach could extend beyond Parkinson’s disease. Other conditions caused by multiple genetic or environmental factors may also benefit from similar classification methods.
“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,” said Verstreken.
No comments yet