Automatic text-mining as an unbiased approach to uncover molecular associations between periodontitis and coronary artery disease

Abstract

Abstract The increasing prevalence of periodontal and cardiovascular diseases is the result of a sedentary lifestyle associated with poor diet, obesity, hypercholesterolaemia, smoking habits, alcohol consumption and stress. The present study aims to uncover molecular associations between periodontitis and coronary heart disease using an unbiased strategy of automatic text mining traditionally applied to bibliometric studies. A total of 1590 articles on these diseases were retrieved from the Web of knowledge database and searched using the VOS viewer to create a network of keywords associated with both diseases. These data were supplemented with data from DisGeNET, which stores known associations to either periodontitis or coronary heart disease. Overall, the automated text mining approach presented here highlighted inflammatory molecules as common associations between periodontitis and coronary heart disease. Specifically, this study showed that molecules such as C-reactive protein, interleukins 6 and 1-β, myeloperoxidase, and matrix metalloproteinase 9 are simultaneously associated with periodontitis and coronary artery disease by both text mining and DisGeNET analyses. This association validates the multiplex assessment of salivary inflammatory markers as a tool to assess cardiovascular disease risk and could become an important tool to identify common molecular targets to monitor both diseases simultaneously. In addition, the text mining protocol and subsequent data processing and methods using bioinformatics tools could be useful to uncover links between other diseases.