Abstract:
|
Cancers are routinely classified into subtypes according to various features, including histo-pathological characteristics and molecular markers. Genomic investigations have reported heterogeneous association between loci and cancer subtypes. However, it is not evident what is the optimal modeling strategy for handling correlated tumor features, missing data, and increased degrees-of-freedom in the underlying tests of associations. We propose score tests for genetic associations using a mixed-effect two-stage polytomous model (MTOP). In the first stage, a standard polytomous model is used to specify for all subtypes defined by the cross-classification of different markers. In the second stage, the subtype-specific case-control odds ratios are specified using a more parsimonious model based on the case-control odds ratio for a baseline subtype, and the case-case parameters associated with tumor markers. Further, to reduce the degrees-of-freedom, we specify case-case parameters for additional exploratory markers using a random-effect model. We use the Expectation-Maximization (EM) algorithm to account for missing data on tumor markers. We derived the two independent score statistics
|