Multi-SNP prediction model for lung function decline

Chronic obstructive pulmonary disease (COPD) is a 3rd leading cause of death (1) which decreases lung function due to irreversible airway obstruction. The main indicator of the progression of COPD is a rate of the forced expiratory volume of 1 second (FEV1) decline. The intern will build the prediction model for the slope of FEV1 decline and find the genetic variants that affect these FEV1 changes. Some variable selection machine learning algorithms will be applied to screen important genetic variants and the performance of prediction on FEV1 change will be compared. The multi-SNP prediction model can be used to classify individuals who are expected to have a rapid decline of FEV1 based on their genetic information. These pre-screened patients can be targeted for frequent medical examinations.

Faculty Supervisor:

Xuekui Zhang

Student:

Songwan Joun

Partner:

Providence Health Care

Discipline:

Statistics / Actuarial sciences

Sector:

Health care and social assistance

University:

University of Victoria