Stable isotope resolved metabolomics classification of prostate cancer cells using hyperpolarized NMR data.

High signal-to-noise metabolic fingerprints can be obtained with hyperpolarized stable isotope resolved 13C-nuclear magnetic resonance. Based on these data machine learning combined with feature selection allow for high precision discrimination of aggressive against indolent prostate cancer cells. The analysis suggests trioses to play a decisive role in the metabolic phenotype of aggressive prostate cancer cells.