Table 4.
VIP | βadj | P-value | FDR | |
---|---|---|---|---|
Nicotinate | 2.27 | 0.11 | 0.009 | 0.20 |
Deoxyinosine | 1.82 | 0.10 | 0.015 | 0.20 |
N-oleoylethanolamine | 1.75 | -0.10 | 0.016 | 0.20 |
Linoleoyl ethanolamide | 1.64 | -0.10 | 0.018 | 0.20 |
Docosapentaenoate | 1.45 | -0.10 | 0.018 | 0.20 |
Xanthine | 2.33 | 0.10 | 0.019 | 0.20 |
Docosapentanoic acid | 1.44 | -0.10 | 0.020 | 0.20 |
Butyrate | 1.83 | 0.10 | 0.020 | 0.20 |
Propionate | 2.52 | 0.09 | 0.025 | 0.22 |
C18:0 Monoacylglycerol | 1.30 | -0.09 | 0.036 | 0.22 |
Eicosatrienoic acid | 1.31 | -0.09 | 0.037 | 0.22 |
Stearoyl ethanolamide | 1.12 | -0.08 | 0.043 | 0.22 |
Glutamate | 1.20 | 0.08 | 0.044 | 0.22 |
Valeric acid | 1.47 | 0.08 | 0.046 | 0.22 |
C16:0 ceramide | 1.16 | -0.08 | 0.046 | 0.22 |
Bilirubin | 1.37 | -0.08 | 0.046 | 0.22 |
Adrenic acid | 1.21 | -0.09 | 0.047 | 0.22 |
VIP refers to the variable importance in projection score provided by PLS. βadj and p-values correspond to the effects of metabolites individually tested in linear regression adjusted for age, BMI, exercise, years since menopause, FSH, and estradiol. Metabolite identities were validated in the original publication for the reference samples used to train the MelonnPan model (see methods).