Project With Haolun Shi

Bayesian Basket Trial Design for a Longitudinal Ordinal Endpoint

In basket trials, subjects all receive the same treatment, which targets a specific mutation or biomarker. Bayesian methods are commonly used for modelling the treatment assignment decisions in these trials. In this work, the student will develop an adaptive method for modelling the time progression term in the latent normal variable of the longitudinal ordinal endpoint. The student will be responsible for:

1.       Conducting a literature review of the competing methods in Bayesian basket trial design

2.       Proposing a Bayesian model for ordinal endpoints

3.       Implement the proposed method in R

4.       Documenting the work in a manuscript