Skip to main content

Table 1 List of the OPLS-DA models that were validated using a cross-validation test with 200 permutations, showing the classes that were used

From: Towards a scientific interpretation of the terroir concept: plasticity of the grape berry metabolome

  

Non-volatile metabolome

Volatile metabolome

OPLS-DA model

Classes

Q2

Q2

Macrozone

Soave vs. valpolicella vs. lake garda

0.541

0.384

Macrozone

Lake garda vineyards vs. others

0.635

0.536

Vineyard altitude

>200 m vs. < 200 m

0.556

0.0117

Row direction

N–S vs. E–W

0.396

0.521

Training system

Parral vs. guyot

-

0.549

Soil type

pH (<8 vs. >8)

0.631

0.621

Soil type

Total lime % (<10 vs. 10–20 vs. >20)

0.671

-

Soil type

Active lime % (<4 vs. > 4)

0.179

0.646

Soil type

Loam % (<30 vs. 30–50 vs. >50)

0.322

0.29

Soil type

Clay % (<25 % vs. >25 %)

0.582

0.226

Soil type

Sand % (<20 vs. 20–50 vs. >50)

0.198

-

Soil type

Organic matter % (<2 vs. 2–2.5 vs. >2.5)

0.588

0.368

Soil type

Exchangeable phosphorus mg/kg (<30 vs. 30–50 vs. >50)

0.174

-

Soil type

Exchangeable potassium mg/kg (<200 vs. 200–400 vs. >400)

0.669

0.426

Soil type

Exchangeable magnesium mg/kg (<300 vs. 300–700 vs. >700)

0.446

0.426

Soil type

Exchangeable calcium mg/kg (3000–8000 vs. >8000)

0.291

0.086

  1. Fully-validated models with a Q2 value greater than 0.5 are shown in bold separately for the non-volatile (LC-MS) and volatile (GC-MS) data sets