Biostatistics PhD student Lucy Gao was awarded a travel scholarship from the American Statistical Association’s (ASA) Biometrics Section to present at the 2019 Joint Statistical Meetings, July 27-August 1, in Denver, Colo. This year's application process was competitive with the ASA Biometrics Section reporting a record-breaking 76 submissions, an increase of 52% from last year.
Gao will present a paper entitled “Are Clusterings of Multiple Data Views Independent?” which she co-authored with Jacob Bien, University of Southern California, and Daniela Witten, University of Washington biostatistics and statistics professor.
The paper is motivated by the Pioneer 100 (P100) Wellness Project which collected many types of data from each of 108 healthy participants at multiple points in time. The goal of the P100 study is to integrate different data types at different times in order to better characterize and optimize wellness. One way to do this is to identify subgroups, or clusters, among the participants on the basis of all of the data types and all of the time points, and then tailor health recommendations for each subgroup. However, this approach assumes that a single underlying clustering of the participants is shared across all data types and all time points. Gao's paper outlines development of a new hypothesis test to evaluate the assumption that there exists some relationship between subgroups defined with respect to different data types and/or different time points. Findings show that while subgroups of the P100 participants defined with respect to any single data type seem to be dependent across time, subgroups among the P100 participants based on one data type (e.g. proteomic data) appears not to be associated with the clustering based on another data type (e.g. clinical lab test data).