Role of genome-wide association studies, polygenic risk score and AI/ML using big data for personalized treatment to the patients with cardiovascular disease

Our partner journal Future Medicine AI has recently published a commentary, which explores the need for portable AI/ML pipelines integrating genomics and healthcare data for discovering novel biomarkers and predicting CVD with high accuracy, ultimately to support clinical diagnostics and decision-making processes.

Introduction:

Big data, GWAS, polygenic risk scores & AI/ML

Big data is a term used to describe extensive and wide data sets that are very intricate, convoluted, multifaceted and cannot be manually analyzed due to human error. Before big data, information was traditionally stored in small copious amounts, making it localized to the individual who inputs these datasets into their system [1]. The principles that define big data can be used in quality improvement under the guise of individual genomic data, such as the sequencing of DNA, RNA and their characteristics, and are therefore tied to clinical decision-making in examples such as personalized medicine [2]. Advancements in technology are driving the growth of personalized medicine, resulting in a parallel expansion of genomic medicine. This expansion aims to enhance individual diagnostics, ultimately leading to a reduction in personal side effects [3,4]. Previously, we reported the significance of an integrated approach that combines gene variant and clinical data [5]. We employed analyses of functional mutations, splice variants, variant distribution and divergence to uncover the importance and prevalence of variants linked to well-studied genes associated with heart failure (HF) and cardiovascular disease (CVD) [5]. Additionally, we have conducted comparative studies where we explored gene identification through multi-ethnic and ancestry-specific studies, how multiple single nucleotide polymorphisms (SNPs) are related to single disease-associated genes, and the correlation of specific biomarkers to both HF and atrial fibrillation (AF) [6]. To advance cardiovascular genomic medicine toward a predictive and preventive paradigm, it is imperative to precisely evaluate disease risk, effectively communicate variant findings, and establish clinical interventions aimed at averting or mitigating the associated ailments. Through a deep understanding of an individual’s entire genome, we can leverage artificial intelligence (AI) and machine learning (ML) models to create a more refined approach for managing patients with CVD.