Catalyst Awardee

Project Description

Detecting Parkinson’s Disease-Associated Metabolites with Colorimetric Sensor Array and Machine Learning

Koh Tong Wey, M.Sc., Ph.D., and Ian Cheong, LL.B., Ph.D., Temasek Life Sciences Laboratory; and Tan Eng Kind, M.B.B.S., National Neuroscience Institute

Competition Sponsor: Ministry of Health and National Research Foundation of Singapore

Parkinson’s disease (PD) is a common aging-related debilitating neurological disorder. No drug can reverse disease progression and clinical diagnosis can be challenging during early stages of the disease. There is an acute need for affordable, objective, biomarker-based methods both for diagnosis of patients and for accurate identification of subjects for new drug trials. We propose the use of colorimetric sensor array and machine learning algorithms to discern changes in urine metabolites as a PD diagnostic. Similar arrays have been designed to mimic sophisticated olfactory sensory systems of animals for discerning complex chemical mixtures. We have validated our array on hepatitis B and various cancers, and the current proposal to test it on PD is novel. Our array consists of 70 hydrogel spots; each spot contains a unique combination of derivatized acrylamides and acrylates that undergo distinct color reactions with urine metabolites. The resulting color pattern of the 70 spots is an emergent property of the urine metabolites, which in turn reflects health status. The color pattern of each array is analyzed by machine learning algorithms for diagnosis. In this pilot project, we will test the array with normal urine and urine spiked with 50 PD-enriched metabolites chosen from literature. Normal and simulated PD urine will be profiled and classified using a pre-trained machine algorithm. The performance will be evaluated in terms of sensitivity, specificity, PPV and NPV. Limit of detection of each metabolite will be assessed. In addition, we have identified colorimetric reactions designed to detect specific metabolites known to be enriched in PD urine and will incorporate these reactions into the array to improve performance. Success of this pilot will lead to a clinical trial for PD diagnostic. We envision that our technology will be generalizable to many aging-related conditions and will enable personalized preventive approaches for healthy longevity.

To learn more about this proposal, email healthylongevity@nas.edu.

Sign up for updates