Dysbiosis of the gut microbiota was identified in individuals with cardiovascular disease (CVD) and modeled using machine learning, according to a study published in Hypertension.
Sequencing data of fecal 16S ribosomal RNA from patients with CVD (n=478) and without CVD (n=473), which were obtained as part of the American Gut Project, were analyzed with 5 supervised machine learning algorithms (random forest, support vector machine, decision tree, elastic net, and neural networks). A total of 39 differential bacterial taxa were identified in patients with vs without CVD.
Using these 39 taxonomic features, machine learning modeling achieved a testing area under the receiver operating characteristic curve (AUC) of approximately 0.58 for random forest and neural networks, 0.57 for elastic net, 0.55 for support vector machine, and 0.51 for decision tree networks.
The investigators trained the machine learning models with the top 500 high-variance features of operational taxonomic units, rather than with bacterial taxa, yielding an improved AUC of approximately 0.65 for random forest. A significant decrease was observed in the AUC and specificity of neural networks (0.48 and 0.46, respectively). The AUC significantly enhanced to approximately 0.70 after limiting the selection to the top 25 highly contributing operational taxonomic unit features.
“These data point to a core set of altered gut microbiota as a common denominator for a variety of clinical presentations of CVD,” noted the researchers.
Aryal S, Alimadadi A, Manandhar I, Joe B, Cheng X. Machine learning strategy for gut microbiome-based diagnostic screening of cardiovascular disease. Hypertension. Published online September 10, 2020. doi: 10.1161/HYPERTENSIONAHA.120.15885
This article originally appeared on The Cardiology Advisor