Table 4.

Relationships between transcriptomic and metabolomic levels

	Dimensions of the principal component analysis of the metabolic level in blood					Dimensions of the principal component analysis of the fatty acids in blood
Modules	Dim1_met	Dim2_met	Dim3_met	Dim4_met	Dim5_met	Dim1_FA	Dim2_FA	Dim3_FA	Dim4_FA	Dim5_FA
Significantly correlated ($Pvalue\le 0.05$)	1	2	8	0	11	2	1	1	1	1
Trend ($Pvalue\le 0.1$)	3	5	6	7	6	4	0	3	3	0

Modules of co-expressed genes were identified from microarray data in the whole blood by using weighted gene correlation network analysis (WGCNA). Circulating biochemical molecules in the plasma were analysed by 1H-NMR (metabolites; met) or target gas chromatography for the lipid fraction subset (FA) and the data were summarized by weighted linear regressions (dim) using principal component analysis (PCA). Correlations were calculated between the module eigengenes (ME) and the first five dimensions of each PCA. The table indicates the number of significant correlations or trends between modules of coexpressed genes and PCA dimensions