Multimodal data integration for enhanced diagnostics and blood pressure regulation in hypertension

Hypertension, a chronic condition affecting over 1 billion people globally, is a leading risk factor for cardiovascular mortality, with nearly half of cases undiagnosed and a third inadequately controlled. The complex pathophysiology of the disease poses significant diagnostic and therapeutic challenges. Our research utilises data from the UK Biobank, a large-scale cohort study of over half a million participants, with a primary hypertension cohort of n=162,261, including 91,214 with poorly controlled hypertension. The first stage of our research investigates rare genetic variants associated with effective blood pressure (BP) regulation in primary hypertension and apparent treatment-resistant hypertension (aTRH). We will identify significant (P < 5 × 10−5) variants with minor allele frequency (MAF) between 0.005-0.01 associated with BP control in both patient cohorts, employing gene burden analysis pipelines to predict the functional consequences of rare but highly impactful variants. The second stage focuses on characterizing effective BP regulation in primary hypertension and aTRH using multimodal data from the UK Biobank. This study will identify novel markers from electronic health records, lifestyle factors, genomics, blood biochemistry, proteomics, and metabolomics that are significantly associated with effective BP control. To integrate findings across these modalities, approaches such as multi-omics factor analysis (MOFA) or similarity network fusion (SNF) will be employed. The final stage aims to develop machine learning (ML) models that integrate multimodal data to predict future hypertension risk more accurately than models based on monomodal data or traditional clinical risk factors. ML interpretability methods will be used for enhancing classification model transparency and to understand which factors are being utilised within the ML models for decision making. This research seeks to enhance diagnostics and therapeutics by identifying novel biomarkers that can inform new treatment strategies or diagnostic models. These advancements aim to overcome current challenges in hypertension management and improve patient outcomes through precision medicine approaches.