Table 5.

Predictive Ability of High-Throughput Omics Approaches to Evaluate the Effects of Glyphosate and MON 52276

Comparison	R2X	R2Y	Q2	pR2Y	pQ2
Glyphosate vs. control
mRNA-seq	0.22	0.94	0.21	0.15	0.07
Small RNA-seq	0.39	0.99	0.52	0.13	0.002
Genome-wide methylation	0.13	0.66	0.03	0.06	0.89
MON 52276 vs. control
mRNA-seq	0.44	0.99	0.64	0.22	0.001
Small RNA-seq	0.34	0.96	0.34	0.15	0.03
Genome-wide methylation	0.12	0.66	0.008	0.72	0.99

OPLS-DA models were performed for each set of omics data. We show estimates of the total explained variation (R2X), variations between the different groups (R2Y), and the average prediction capability (Q2). We assessed the significance of our classification using permutation tests. New estimates of R2Y and Q2 values were calculated from this 1000 times permuted dataset (p-values pR2Y and pQ2 for permuted R2Y and Q2, respectively).