清华合成与系统生物学中心
Tsinghua Center for Synthetic and Systems Biology

Unsupervised phenotyping of Severe Asthma Research Program participants using expanded lung data

Time: Dec.17 (next Tuesday) 10:30-11:30am
Place: FIT 1-312
Speaker: Dr. Wei Wu
Title: Unsupervised phenotyping of Severe Asthma Research Program participants using expanded lung data
Abstract:
Previous clinical studies have identified asthma phenotypes based on small numbers of clinical, physiologic or inflammatory characteristics. However, no studies have utilized a wide range of variables using machine learning approaches.
In this work, our goal is to identifysubphenotypes of asthma utilizing blood, bronchoscopic, exhaled nitric oxide and clinical data from the Severe Asthma Research Program using unsupervised clustering, and then characterize them using supervised learning approaches. In order to achieve this goal, we appliedunsupervised clustering approaches to 112 clinical, physiologic and inflammatory variables from 378 subjects. Variable selection and supervised learning techniques were employed to select relevant and nonredundant variables, address their predictive values, as well as the predictive value of the full variable set.
We identified ten variable clusters and six subject clusters, which differed and overlapped with previous clusters identified by clinicians. Traditionally defined severe asthmatics distributed through subject Clusters 3-6. Cluster 4 identified early onset allergic asthmatics with low lung function and eosinophilic inflammation. Later onset, mostly severe asthmatics with nasal polyps and eosinophilia characterized Cluster 5. Cluster 6 asthmatics manifested persistent inflammation in blood and bronchoalveolar lavageand exacerbations despite high systemic corticosteroiduse and side effects. Age of asthma onset, quality of life, symptoms, medications and health care utilization were some of the 51 nonredundant variables distinguishing subject clusters. These 51 variables classified test cases with 88% accuracy, compared to 93% accuracy with all 112 variables.
The machine learning approaches used here provide unique insights into asthma pathogenesis while revealing novel but readily clinically recognizable phenotypes.

Short Bio:
Dr. Wei Wu is an Associate Research Professor at the Lane Center for Computational Biology at Carnegie Mellon University. Her research focuses on understanding complex human diseases by undertaking integrative approaches, which combine biology, computational and statistical learning, bioinformatics, and genomics. Dr. Wu received a Ph.D. in Computational Molecular Biology from Rutgers University. She also received a M.S. in Computer Science from the University of California at Santa Cruz, where she worked on the construction of the Human Genome Browser with Professor David Haussler. She later did a postdoctoral training at Lawrence Berkeley National Lab with Dr. Mina Bissell. Her current work involves: i) subphenotyping asthma patients using computational approaches; and ii) understanding breast cancer mechanisms using dynamic network learning approaches. Her paper on developing a tree-varying network learning approach for analyzing a breast cancer progression series of cells has won the Best Paper Award at the ISMB conference in 2012. She is a co-Principle Investigator on two NIH R01 awards.