Advisor(s)
Nengjun Yi
Committee Member(s)
Fazlur Rahman
Peng Li
Satwik Acharyya
Xiaoxiao Zhou
School
School of Public Health
Document Type
Dissertation
Department (new version)
Biostatistics
Date of Award
9-11-2025
Abstract
Survival analysis integrating microbiome and clinical data offers powerful insights into human health. The microbiome influences immunity, inflammation, and cancer outcomes. Combining microbial profiles with clinical factors like tumor stage, treatment, and demographics enhances understanding of how host-microbe interactions affect survival. To analyze these complex datasets, advanced statistical methods, including Bayesian models, penalized Cox proportional hazards models, and machine learning techniques are employed. These approaches can uncover novel biomarkers and therapeutic targets, leading to personalized treatment strategies that optimize patient survival. However, challenges arise due to the compositionality, high-dimensionality, and phylogenetic relation between taxa in microbiome data. In this dissertation, in the first part, we propose a compositional Bayesian Cox proportional hazards model (BCCox PH) with a regularized horseshoe prior to analyze compositional microbiome and clinical data. We apply a soft sum-to-zero constraint to the microbiome variable to deal with compositionality. Also, we consider the phylogenetic structure of the microbiome and introduce a structured shrinkage prior incorporating the similarity to deal with it. In the second part, based on the natural hierarchy levels of microbiome data, we add sub-compositional constraints to our model and check the model predictive performance. To evaluate the predictive performance of our model, we conducted extensive simulation studies and an appropriate real data analysis. In the third part, we develop a comprehensive R package based on our proposed models, which predicts multiple types of outcomes (e.g. continuous, binary, survival, et.al) for clinical and compositional microbiome data. The implementation is carried out using the R package brms, with results summarized based on two Markov Chain Monte Carlo (MCMC) algorithms executed in Stan.
ProQuest ID
Recommended Citation
Ding, Zhenying, "Bayesian Survival Analysis For High-Dimensional Compositional Data" (2025). All ETDs from UAB. 7376.
https://digitalcommons.library.uab.edu/etd-collection/7376
