Table 6. Predictive power of principal component analysis, weighted network analysis and random forest regression in drip loss, pH1, pH24 and meat color based on a multiple regression model.

trait	multiple correlation coefficients
	10 principal components of PCA			10 modules of WNA			10 metabolites with highest variable importance of RFR
	RMSE	R2	CV[%]	RMSE	R2	CV[%]	RMSE	R2	CV[%]
drip loss	1.13	0.07	59.75	1.10	0.18	57.94	1.10	0.32	58.13
pH1	0.43	0.35	6.53	0.44	0.30	6.69	0.43	0.37	6.64
pH24	0.32	0.27	5.73	0.32	0.27	5.78	0.34	0.12	6.13
color	2.54	0.23	3.50	2.56	0.21	3.53	2.51	0.37	3.46

PCA = principal component analysis; WNA = weighted network analysis; RFR = random forest regression; RMSE—root mean square error; R²—coefficient of determination; CV—coefficient of variation; drip loss measured in Musculus longissimus dorsi (LD) 24 h post-mortem (p.m.); pH1 measured in LD 45 minutes p.m.; pH24 measured in LD 24 h p.m.; color = meat color measured in LD 24 h p.m.