Skip to main content
. 2019 Jul 10;14(7):e0219403. doi: 10.1371/journal.pone.0219403

Table 4. OPLS analysis of metabolites with variable importance in projection for urinary protein levels.

A. 12 weeks
Positive predicting variable Negative predicting variable
Metabolites VIP Metabolites VIP
Coniferyl alcohol-2TMS 0.0864 ± 0.0678 Glycine-3TMS -0.1486 ± 0.0766
Proline-2TMS -0.1460 ± 0.0796
Methionine-2TMS -0.1445 ± 0.0439
Sorbose-meto-5TMS(1) -0.1410 ± 0.0542
Thiodiglycolic acid-2TMS -0.1392 ± 0.0551
Psicose-meto-5TMS(2) -0.1387 ± 0.0411
B. 22 weeks
Positive predicting variable Negative predicting variable
Metabolites VIP Metabolites VIP
Fucose-4TMS(1) 0.1386 ± 0.0951 Gluconic acid-6TMS -0.1638 ± 0.0937
Galacturonic acid-5TMS(1) 0.1029 ± 0.0949 Glyceraldehyde-meto-2TMS(2) -0.1426 ± 0.1055
Lysine-4TMS -0.1389 ± 0.0643
Glyoxylic acid-meto-TMS -0.1387 ± 0.1020
Isoleucine-2TMS -0.1324 ± 0.1177
allo-Isoleucine-2TMS -0.1324 ± 0.1177
4-Aminobutyric acid-2TMS -0.1324 ± 0.1177
Thiodiglycolic acid-2TMS -0.1283 ± 0.1037
C. 32 weeks
Positive predicting variable Negative predicting variable
Metabolites VIP Metabolites VIP
Aconitic acid-3TMS 0.1524 ± 0.0705 Serine-3TMS -0.1338 ± 0.1136
Myristic acid-TMS 0.1188 ± 0.0505 Propionylglycine-TMS -0.1330 ± 0.1247
Citric acid-d4-4TMS 0.1073 ± 0.1070 4-Hydroxybutyric acid-2TMS -0.1082 ± 0.1046
Rhamnose-4TMS(1) -0.0966 ± 0.0859

Data are the mean ± SEM. VIP means variable importance in projection, which was calculated in OPLS analyses using SIMCA.