Innovative training in methods for future data

Annah Muli

I am from Kenya, a beautiful country in East Africa with coastline on the Indian Ocean and encompasses the savannah, lakelands, the dramatic Great Rift Valley and mountain highlands. My graduate academic journey began at the University of Nairobi where I did my bachelor’s degree in statistics and a master’s degree in mathematical statistics.

I was fortunate to join the IMforFUTURE network at the University of Leeds on August 2018 under the supervision of Prof. Jeanine Houwing-Duistermaat and Dr. Arief Gusnanto. My research focuses on developing flexible statistical methods for analysis of (correlated) survival data that will be applied to omics data. Survival data entails data about time to the occurrence of an event and is widely applicable across many disciplines. Overtime, there has been a high reliance on the Cox proportional hazards model for modelling of survival data. However, the Cox model’s assumptions are often violated in omics research which necessitates the need for alternative flexible models.

The study design will involve adjustments for correlation, delayed entry and ascertainment. A method based on inverse probability weighting has already been developed but it assumes a parametric hazard. We seek to extend this method to semi-parametric Cox models by using splines for the hazard and consider flexible models for covariates.