【讲座】Metabolomics to Discover Biomarkers for TB Diagnosis - A Statistician's View
2019.07.09 活动

You are cordially invited to the iDDA Academic Seminar, to be delivered by Professor Myung Hee LEE from Weill Cornell Medical College at 14:00 - 15:00 on July 23, 2019 (Tuesday).

 

Title:  Metabolomics to Discover Biomarkers for TB Diagnosis - A Statistician's View

Time & Date: 14:00 - 15:00, July 23, 2019 (Tuesday)

Venue: Room 208, Cheng Dao Building

Speaker: Professor Myung Hee LEE, Weill Cornell Medical College

 

 

Abstract

 

Variable selection for high dimensional data analysis plays a pivotal role in many fields of modern science. I will introduce the problem of variable selection arising from binary-class high dimensional data. The dataset consists of a large number of metabolites signatures, simultaneously measured from urinary samples from tuberculosis (TB) patients and healthy controls, where severe batch effect is present. I will present an iterative empirical Bayes approach for feature discovery with an explicit goal of capturing unwanted variation such as batch effects. Then, I will describe unique challenges (and opportunities) arising from analyzing such metabolites datasets. Time permitting, I will also present a brief overview of some additional statistical projects being done with collaborators at Weill Cornell Medicine in NYC.

 

 

Biography

 

cid:image003.jpg@01D5356E.D1554A00

Myung Hee Lee is an Assistant Professor of Statistics in Clinical Epidemiology at Weill Cornell Medical College. She has a broad background in Statistics, with specific training and expertise in high dimensional, low sample data analysis. She received her Ph.D. from the University of North Carolina at Chapel Hill. She joined the bioinformatics group at the Lineberger Comprehensive Cancer Center, UNC-CH in 2007 shortly after graduation, and was a member of statistics faculty at Colorado State University from 2008-2015. The primary focus of her research is to develop statistical methodologies for high dimensional, low sample size data. Her statistical work has been largely motivated by biological applications, particularly in analyzing high throughput data where the number of variables exceeds the sample size such as transcriptomics and metabolomics.

 

She is a faculty member in the Weill Cornell Center for Global Health where she is the primary statistician for the research projects. She also mentor clinical epidemiology graduate students, post docs, and junior faculty conducting research at her Center’s international sites (Brazil, Haiti, India, Tanzania).

 

 

 

Institute for Data and Decision Analytics (iDDA)

The Chinese University of Hong Kong, Shenzhen

cid:image018.jpg@01D530BF.24AE57F0