Introduction
A number of genetic variants are associated with risk of hypertension and expression quantitative traits, there is, however, a lack of insights into complex biological mechanisms of hypertension. This project aims to develop a machine learning approach to explore the molecular causal networks of hypertension, using data from large-scale genome-wide association studies of blood pressure, and measurements of gene expression, splicing and DNA methylation from human kidneys. This work will help target diagnostic biomarkers and therapeutic treatments for hypertensive and cardiovascular medicine.
Explaining the science
High blood pressure is a leading cause of premature death worldwide. Many factors (for example, genes) may drive the risk of high blood pressure themselves, or through other risk factors. In order to target the right factors for the development of effective diagnosis and treatment, it is important to understand how these risk factors are related to each other and how they work together.
Kidney is an important organ to study these relationships. However, there is a lack of robust yet flexible statistical methods for learning the complex biological structure underlying high blood pressure in the literature. In this project, a machine learning method will be developed to learn causal networks of high blood pressure from a unique dataset including biological measurements collected from human kidneys, together with summery data from large-scale genome-wide association studies.
Project aims
Blood pressure associated genetic variants may lead to changes in blood pressure and the risk of hypertension through activating chains of molecular events at various levels of transcriptional regulation within human kidney. This project aims to
1) identify blood pressure associated genetic variants that overlap with expression QTLs (eQTLs), splicing QTLs (sQTLs) and methylation QTLs (mQTLs);
2) build directed molecular causal network of hypertension.
A machine learning approach will be developed to explore causal mechanisms underlying hypertension by including multiple potential risk factors at multiple molecular levels simultaneously in analysis.
This work will help gain biological insights into pathogenesis of hypertension, and target more precisely diagnostic biomarkers and therapeutic treatments for hypertensive and cardiovascular medicine. The output of this project will be of direct interest to clinicians and researchers in cardiology for the benefit of patients.
Applications
This work can be applied to clinical research in cardiology to focus attention on candidate causal biomarkers of hypertension. The analytic framework developed in this project will be applicable to studies of a wide range of clinical outcomes of interest, and thus in many health-related applications.