Statistical inference regarding within-family association in disease onset times

Speaker: Richard Cook, Professor, University of Waterloo, Canada.


In studies of the genetic basis for chronic conditions, interest routinely  lies in the within-family dependence in disease status.  When probands are selected from  disease registries and their respective families are recruited, a variety of  ascertainment  bias-corrected methods of inference are available which are typically based on models for  correlated binary data. This approach inadequately deals with the ages family members are  at the time of assessment and hence the variation in the time at risk among participants. We consider copula-based models for assessing the within-family dependence in the  disease onset time and disease progression, based on right-censored and current status  observation of the onset times for non-probands. These models are also used to examine the  factors influencing the commonly used measures of within-family dependence for correlated  binary response to address parent of origin hypotheses. The methods are applied to data from  a family study at the University of Toronto Psoriatic Arthritis Clinic.

Joint work with Yujie Zhong, University of Waterloo.

Published Aug. 13, 2015 5:19 PM