Multiomics Analysis Coupled with Text Mining Identify Novel Biomarker Candidates for Recurrent Cardiovascular Events

Abstract

Recurrent cardiovascular events remain an enigma that accounts for >30% of deaths worldwide. While heredity and human genetics variation play a key role, host–environment interactions offer a sound conceptual framework to dissect the molecular basis of recurrent cardiovascular events from genes and proteins to metabolites, thus accounting for environmental contributions as well. We report here a multiomics systems science approach so as to map interindividual variability in susceptibility to recurrent cardiovascular events. First, we performed data and text mining through a mixed-methods content analysis to select genomic variants, 10 single nucleotide polymorphisms, and microRNAs (miR-10a, miR-21, and miR-20a), minimizing bias in candidate marker selection. Next, we validated our in silico data in a patient cohort suffering from recurrent cardiovascular events (a cross-sectional study design and sampling). Our findings report a key role in low-density lipoprotein clearance for rs11206510 (p < 0.01) and rs515135 (p < 0.05). miR-10a (p < 0.05) was significantly associated with heart failure, while increased expression levels for miR-21 and miR-20a associated with atherosclerosis. In addition, liquid chromatography-mass spectrometry-based (LC-MS-based) proteomics analyses identified that vascular diameter and cholesterol levels are among the key factors to be considered in recurrent cardiovascular events. From a methodology innovation standpoint, this study offers a strategy to enhance the signal-to-noise ratios in mapping novel biomarker candidates wherein each research and conceptual step were interrogated for their validity and in turn, enriched one another, ideally translating information growth to knowledge growth.