This article was originally published here
J Biomed Inform. 2021, April 28: 103796. doi: 10.1016 / j.jbi.2021.103796. Online before printing.
ABSTRACT
Individual variations in genetic and environmental factors can cause differences in metabolic phenotypes that can affect patient drug responses. In-depth study of patient responses to therapeutics is a critical and urgent event in the personalized treatment study. Using machine learning methods to discover aptitude assessment biomarkers can provide deep insight into the mechanism of disease therapy and facilitate the development of personalized medicine. To find important metabolic network signals for predicting patient drug responses, a novel method called differential metabolic network construction (DMNC) has been proposed. In DMNC, changes in the concentration of metabolite ratios between different pathological conditions are measured to create different metabolic networks that can be used to aid clinical decision-making. In this study, DMNC was used to characterize the responses of patients with type 2 diabetes mellitus (T2DM) to modified-release gliclazide therapy (MR). Two T2DM metabolomics datasets from different batches of subjects treated with gliclazide MR were analyzed in detail. A network biomarker was defined to assess patient eligibility for gliclazide MR. It can be effective in predicting significant responders from insignificant responders and can achieve an area below the curve values of 0.893 and 1000, respectively, for the detection and validation sets, respectively. Compared to the metabolites selected by the other methods, the network biomarker selected by DMNC was more stable and precise to reflect the metabolic responses in patients to gliclazide MR therapy, thus contributing to the personalized medicine of T2DM patients. The better performance of DMNC confirmed its potential to identify network biomarkers to characterize responses to therapeutic treatments and provide valuable information for personalized medicine.
PMID: 33932596 | DOI: 10.1016 / j.jbi.2021.103796
