Brain research and advocacy non-profit Cohen Veterans Bioscience (CVB) announces the publication of results from its data sciences research program.  As part of the drive toward precision medicine, there has been an increased focus on the discovery of biological markers and quantitative techniques to serve as diagnostic and prognostic tools for individual patients, and for monitoring the progression or remission of disease. 

Published in Frontiers In Aging Neuroscience on February 13, 2023, the article, titled “Machine learning within the Parkinson’s progression markers initiative: Review of the current state of affairs” presents results from analysis of more than a decade’s worth of longitudinal and multi-modal data from patients, healthy controls, and at-risk individuals, including imaging, clinical, cognitive, and ‘omics’ biospecimens. Led by Cohen Veterans Bioscience with support from The Michael J. Fox Foundation, it provides an overview of the application of machine learning methods to analyzing data from the PPMI cohort.

The review explores the potential to combine multiple modalities to gain a broader perspective on biological mechanisms of patient heterogeneity while uncovering a striking lack of overlap between findings across studies.

“The pioneering PPMI study has generated a unique and rich dataset that will shed light on pathological pathways, subtypes and progression of Parkinson’s Disease. This work was a critical first step in our mission to discover and develop tests and treatments for neurodegenerative diseases”, Lee Lancashire, Principal Investigator of the study and Chief Information Officer at CVB “With Artificial Intelligence (AI) playing a more prominent role in biomedical research, we wanted to understand the existing gaps and untapped resources that would then enable us to make recommendations to the field that may guide future discovery efforts that strive to unlock better outcomes across these complex neurological diseases.” 

Parkinson’s disease is the second most common neurodegenerative disease and its prevalence has been projected to double over the next 30 years. While accurate diagnosis of Parkinson’s disease remains challenging, the field is evolving from a clinical to a biomarker-supported diagnostic entity, for which earlier identification is possible, different subtypes with diverse prognosis are recognized, and novel disease-modifying treatments are in development.

“To further advance this promise of precision therapeutics we need more investment in multimodal modeling of the brain”, said Dr. Magali Haas, CEO & President at CVB.  “Most brain diseases are multifactorial, complex conditions requiring us to rely on machine-learning approaches to build models that capture individual variability in genetics, pathways, circuits and systems.”

This work was jointly funded by Cohen Veterans Bioscience (COH-0003) and a generous grant from the Michael J. Fox Foundation as part of the Parkinson’s Progression Markers Initiative (PPMI) (MJFF-019075). Data used in the preparation of this article were obtained from the Parkinson’s Progression Markers Initiative (PPMI) database (www.ppmiinfo.org/data).

Gerraty, R.; Provost, A.; Lin, L.; Wagner, E.; Haas, M.; Lancashire, L. Machine learning within the Parkinson’s progression markers initiative: Review of the current state of affairs. Frontiers In Aging Neuroscience 13 February 2023 https://doi.org/10.3389/fnagi.2023.1076657