Table 1.
Correlation between all standardized metabolite datasets investigated in this study.
| Data source | Pairwise conditions | Days |
R of standardized data |
||||
|---|---|---|---|---|---|---|---|
| Glucose | Lactate | Glutamine | Glutamate | Ammonia | |||
| Off-line | CuBiAn/LC-MS | 1–9 | 0.968 | 0.979 | 0.944 | 0.777 | N/A |
| Chemometrics | Chemometrics/LC-MS | 1–9 | 0.975 | 0.983 | 0.937 | 0.766 | N/A |
| Chemometrics/CuBiAn | 1–10 | 0.987 | 0.986 | 0.922 | 0.829 | 0.936 | |
| Raman univariate predictors | Best positively correlated/LCMS | 1–9 | 0.971 | 0.952 | 0.839 | 0.845 | N/A |
| Best positively correlated/LCMS | 2–9 | 0.970 | 0.944 | 0.926 | 0.835 | N/A | |
| Best positively correlated/CuBiAn | 1–10 | 0.958 | 0.673 | 0.654 | 0.177 | 0.935 | |
| Best positively correlated/CuBiAn | 2–10 | 0.956 | 0.701 | 0.899 | 0.173 | 0.931 | |
| Overall means | 0.97 | 0.89 | 0.87 | 0.63 | 0.93 | ||
The datasets consisted of off-line CuBiAn bioanalyzer data for glucose, lactate, glutamine, glutamate, and ammonia collected daily over 10 days.
Bold numbers indicate the strongest correlation among all pairwise analyses for each analyte. The underlines numbers indicate the highest correlations obtained using the blind univariate approach.