Skip to main content

Table 5 Prediction of summed saturated fatty acids (SAT) content in yeast biomass by FTIR spectroscopy, using partial least square regression (PLSR) analysis

From: Biochemical profiling, prediction of total lipid content and fatty acid profile in oleaginous yeasts by FTIR spectroscopy

Spectral region (cm−1)

\({{\text{RMSE}}_{{{\text{CV}}}}} ^{{\text{a}}}\)

(N = 125)

RMSETest

(N = 56)

\({R^{2} {_{{{\text{CV}}}}}} ^{{\text{b}}}\)

(N = 125)

R 2 Test

(N = 56)

NFactorsc

3100–2800

4.50

6.70

0.78

0.73

11

3100–2800. 1800–700

3.97

12.74

0.83

0.15

14

3100–2800. 1800–1700

5.46

8.06

0.67

0.50

4

3100–2800. 1700–1715

4.64

5.68

0.77

0.70

12

3100–2800. 1800–1715

4.41

5.70

0.79

0.69

17

1800–1700

6.91

13.28

0.51

0.02

5

1800–700

5.51

11.58

0.67

0.23

5

  1. N is the number of samples in the dataset
  2. aRMSE—root mean squared error of cross-validation in calibration (RMSECV) and test set (RMSETest)
  3. bR2—coefficient of determination for cross-validation in calibration (R2CV) and test set (R2Test)
  4. cNFactors—number of factors used in PLSR models