Skip to main content
Fig. 3 | BMC Bioinformatics

Fig. 3

From: A generalized covariate-adjusted top-scoring pair algorithm with applications to diabetic kidney disease stage classification in the Chronic Renal Insufficiency Cohort (CRIC) Study

Fig. 3

Top-scoring pairs for discriminating DKD stage from among the raw metabolite ions. a Scatter plot for the top pair of raw metabolite ions selected by the TSP algorithm along with TSP’s decision boundary. The axes are metabolite ion abundances that were creatinine normalized and natural log transformed. Patients had either early-stage DKD (stage G2-3b, N = 777) or advanced-stage DKD (stage G4, N = 200). b Heatmap correlation matrix of clinical variables vs raw metabolite ions selected by the K-TSP algorithm. Single ion can annotate to multiple metabolites, which resulted in ambiguity in assignments. Metabolite ion (full annotated name): Ion 13: 3,6-Dihydro-4-(4-methyl-3-pentenyl)-1,2-dithiin. Ion 17: 13,14,15-trihydroxy-9-oxo-8,17-dioxatetracyclo[8.7.0.02,7.011,16]heptadeca-1(10),2(7),3,5,11,13,15-heptaen-5-yl acetate. Ion 19: 2-Phenylethyl beta-d-glucopyranoside

Back to article page

BMC Bioinformatics

ISSN: 1471-2105

Contact us