Leveraging Association Rule Mining to Detect Pathophysiological Mechanisms of Chronic Kidney Disease Complicated by Metabolic Syndrome

The purpose of this study is to explore pathophysiological mechanisms present in patients that suffer from Chronic Kidney Disease complicated by Metabolic Syndrome (CKD-MetS) so as to better support proactive treatment. Association rule mining was applied to extract significant associations from the Semantic MEDLINE Database (SemMedDB). A total of 23,310 PMIDs with 5,542 unique items were included in our dataset. We focused on 5 specific syndromes that were extracted: diabetes, cardiovascular disease, increased triglycerides, obesity and inflammation. The number of rules generated for these five diseases are SO, 197, 31, 21 and 21 respectively. Our study identified several pathophysiological mechanisms that exist in CKD-MetS patients that can contribute to further renal damage.