Improving the analysis of designed studies by combining statistical modelling with study design information

BMC Bioinformatics

Table 1 Summary of the regression models

Model nr	Effects included for both the metabolites and for SAA	Q²_CV	PLS factors	Metabolites: Used variation (%)	SAA: Used variation (%)
1	All data	0.52	6	100	100
2	Residual	0.44	4	63	41
3	Residual, Time	0.51	4	75	45
4	Residual, Diet	0.63	5	76	89
5	Residual, Interaction	0.56	4	74	48
6	Residual, Time, Diet	0.63	6	89	93
7	Residual, Time, Interaction	0.56	5	86	52
8	Residual, Diet, Interaction	0.58	6	87	96

The Q²_CV values are presented for the predicted SAA values and the measured ones (n = 61). The prediction of the SAA values was done with double 12-fold cross-validation. The Q²_CV value represents the percentage of variation of SAA that is explained by the variation in metabolites.

ISSN: 1471-2105