Figure 2.

Distinguishing blinded “left out” sera samples, and LOOCV with a lower cost desk top mass spectrometer. (A) Six blinded epilepsy and 4 control samples from each group were removed from the database and the remaining samples (N = 23 for epilepsy and 13 for control) were used as a training set. A % LOOCV classified mass peak group test metric cut-off of 43.0% was established 3.26 SD’s from each respective group mean as described in the methods. (B) Six blinded epilepsy samples and 4 control samples were LOOCV tested against the test metric cut off of 43.0% determined by the training group dataset. (C) Epilepsy patient and control % scores for the LOOCV dataset (N = 29 epilepsy and 17 control) obtained using a small-footprint, lower ion-resolution, and limited m/z range (200–1200) MS instrument (Advion, Inc.). Group test metric cut off is 35.16% with 1.22 SD. (D) Random LOOCV grouping of epilepsy patients and controls from panel C demonstrates lack of group separation (p value 0.37). A lower resolution and lower cost MS instrument than the LCQ Advantage (Expression CMS ESI-MS instrument, Advion, Inc.) performed reasonably well in distinguishing subjects of the epilepsy group (N = 29) as different from the control individual (N = 17) group (Figure 2, panels C,D). The mass analyzers in these two instruments have differing physics and electronics (ion trap vs. single quadrupole), and the group discriminatory p-value is larger for the CMS instrument (10⁻¹⁰) than the LCQ instrument (10⁻²²). Whereas the LCQ had no false positives or false negatives, the CMS instrument resulted in three false positives and three false negatives, Figure 2C. The randomization p-value is higher with the CMS instrument vs. the LCQ instrument (0.37 vs. 0.10). These results do suggest a less accurate instrument with reduced m/z range can still detect enough mass spectral signal differences between these two groups. This ability strengthens the conclusions that the biomolecules observable in the serum and differing among the study groups could possibly help in the diagnosis and monitoring epilepsy.