Figure 1. Schematic diagram of the LC-MS based metabolomic workflow.

Wherein the multi-region serum and plasma samples are extracted and analyzed within a single sequence on the ESI-TOF in positive mode. Mass spectral data is preprocessed with XCMS software and multivariate statistical analysis and machine learning classification algorithms are used to distinguish patterns in the data and provide a binary output to the classification of samples. ROC curves are used to quantify the relationship between sensitivity and specificity for a given test. Ultimately, this information can be used in an iterative fashion to interrogate larger datasets and provide necessary diagnostic information to better characterize the disease status of clinical samples from a variety of geographic regions.