Statistics Toolbox for Study: ST001706

Title: Machine learning-enabled renal cell carcinoma status prediction using multi-platform urine-based metabolomics NMR (part-II)

Select a dataset:

NMR data

Run analyses on data in Study ST001706 Dataset: NMR data

Metabolite structures and classes (all analyses combined)

• Pie chart of metabolite super classes
• Pie chart of metabolite main classes
• Pie chart of metabolite sub classes
• Molecular structure similarity analysis

Normalization and averaging

• Perform sample normalization / Show metabolite averages / Run cluster analysis
• Perform analyte scaling on data
• Create Relative log abundance plots

Univariate analysis

• Perform multi-condition dot plot analysis
• Perform Volcano plot analysis
• Perform ANOVA analysis

Clustering and correlation

• Perform hierarchial or heatmap cluster analysis
• Perform Clustered correlation analysis
• Perform Network analysis on correlated metabolites (mapped to classification)
• Perform Network analysis on correlated metabolites (mapped to fold-change)

Multivariate analysis

• Perform Principal component analysis
• Perform Linear discriminant analysis
• Perform Partial least-squares discriminant analysis (PLS-DA)

Classification and feature analysis

• Perform OPLS-DA and VIP projection
• Random Forest and VIP projection

MetaBatch Omic Browser (MD Anderson Cancer Center)

(Clustered Heat Maps, PCA+, UMAP, box plot, violin plot, and other visualizations)

• Load this study
• Load this analysis (AN002779)

MetENP: Metabolite enrichment and species-specific pathway annotation

• MetENP R package
• MetENP tutorial

Mapping metabolites to human biochemical pathways

• Map study metabolites to HMDB and KEGG pathways
• Map study metabolites to pathways with ratio/t-test data