The goal of our project is to calculate biological age using questionnaire responses and proteomic data from buccal swabs collected from our study participants. Determining biological age is significant because it provides a more accurate measure of an individual’s health and aging process than chronological age, allowing for personalized interventions to improve longevity. It also helps identify early signs of age-related diseases, enabling proactive healthcare and lifestyle adjustments.
So far, we have developed explainable machine learning models for predicting biological age using data from NHANES and are looking to validate our model using a combination of proteomic and questionnaire data from the UK Biobank. The model will produce a biological age from the questionnaire data which we look to align with ageing mechanisms from the proteomic data from non-invasive buccal swab testing. Further, UK Biobank data will serve to bridge the NHANES dataset and our in-house proteomics data in hopes to enhance the accuracy and explainability of biological age predictions. Incorporating SHAP in the machine learning models enables us to break down the factors impacting the biological age prediction. Our plan is to provide the generated data and results freely to the wide public, where users can monitor their biological age and observe how their lifestyles choices impact their health.