Abstract:
|
In the Big Data era, a huge amount and variety of data, many of these with real time, from different domains are collected and aggregated in order to address critical scientific questions or improve business performance. For example, the NIH newly-developed Precision Medicine Initiative-Cohort Program (PMI-CP) is planning to collect data from 1 million or more individuals, which include the data of demographics, self-reported measures (symptoms, quality of life, smoking, alcohol), behavioral and lifestyle (diet, physical activity), sensor-based and wearable device data (location, activity monitors), clinical/phenotype data (EMR, imaging, EEG), omics, molecular and biomarker data, environmental factors (GPS, air quality, climate) and social network data. Although many of these data are snapshots of clinical outcomes and risk factors, many others are time course and functional data (including imaging, EEG, location-based data etc.). It is quite challenging to integrate all these different types of data to address scientific questions. I will discuss our recent work in this direction and some new ideas on how to deal with high-dimensional functional data in precision medicine.
|