Title: Quantifying wellness and identifying disease transitions with personal, dense, dynamic, data clouds
Advisors: Larry Ruzzo and Nathan Price
Supervisory Committee: Larry Ruzzo (Co-Chair), Nathan Price (Co-Chair, BioE), Daniela Witten (GSR, STAT), and Ed Lazowska
Abstract:
Personal, dense, dynamic, data clouds (pD3 clouds); where diverse measures from genetics, protemics, clinical labs, metabolomics, microbiome, and personal health devices are taken longitudinally in large cohorts; open a window to individual health that has not existed on this scale before. Computational analysis of such rich health data on individuals presents the opportunity to build a 21st century learning healthcare system, that focuses not only on treating disease, but on promoting wellness and the prevention of disease.
Three major challenges presented by this unprecedented wealth of personal health data is the development of measures and techniques for: 1) understanding an individual's current wellness status; 2) personalized prediction of the effect of interventions on that wellness status; and 3) identifying early transitions to disease states, when that disease is most treatable. My work is addressing these challenges through application of statistical modeling and machine learning to pD3 clouds.
This thesis proposal will present my work in the larger context of Precision (P4) Medicine; discuss prior attempts to use "biological age" as a holistic wellness measure and my current progress developing this measure using pD3 clouds; and present a new technique for developing systems biomarkers to identify the earliest signs of disease transitions. I will demonstrate the feasibility of this proposed thesis by presenting preliminary work using pD3 clouds shared for research purposes by thousands of individuals enrolled in the Arivale program, a pD3 cloud based wellness start-up spun from the Institute for Systems Biology.