Zhiyue Huang

In Beigene, I am responsible to contribute and review protocol, write statistical analysis plan, provide consultation to medical colleagues, analyze data and develop statistical methodologies. I am involved into clinical trials in a wide range of phases. Furthermore, I support medical affairs for publications and health economics and outcome research. My projects are mainly in the area of hematology and lung cancer. My current research interest is on statistical inference on time-to-event data after propensity score matching and indirect comparison.

In the MRC Biostatistics Unit, I joined the European Prevention of Alzheimer's Dementia Consortium for developing novel statistical methodologies for Alzheimer's and dementia data. My major research contribution include, detecting treatment effects with multiple endpoints in longitudinal clinical trials for Alzheimer's disease, non-linear modelling for cognitive scores with the presence of practice effects, and validating Jack's hypothetical pathological model for Alzheimer's disease.

During my Ph.D., I focus on non-parametric mixture/mixed effects models, in which the mixing parameter/random effects distribution is non-parametric. Using reparameterization and approximation techniques, I reduce the infinite dimensional parameter space into a finite one, but with boundaries. Further, I use ensemble techniques to achieve the asymptotic normality for the inference on regression parameters in mixed effects models. My method can be further applied to stratified medicine.