Multiblock variable influence on orthogonal projections (MB-VIOP) for enhanced interpretation of total, global, local and unique variations in OnPLS models

Table 5 Summary of the number of variables used for the block-sPLS models (the original and the two reduced models) and the percentages of explained total variation for the Hybrid Aspen data

Data	Block-sPLS models	Number of variables used	Explained total variation (%)
Transcript	Original block-sPLS	14,738	68.0
	Block-sPLS comparable to MB-VIOPtot ≥ 0.5 model	13,151	68.0
	Block-sPLS comparable to MB-VIOPtot ≥ 1.0 model	4483	66.0
Protein	Original block-sPLS	3132	50.0
	Block-sPLS comparable to MB-VIOPtot ≥ 0.5 model	2201	50.0
	Block-sPLS comparable to MB-VIOPtot ≥ 1.0 model	685	48.0
Metabolite	Original block-sPLS	281	54.0
	Block-sPLS comparable to MB-VIOPtot ≥ 0.5 model	236	54.0
	Block-sPLS comparable to MB-VIOPtot ≥ 1.0 model	77	52.0

The information has been distributed in three areas according to data block (transcriptomics, proteomics and metabolomics), and each area is divided in three rows: one for the original model, one for the reduced model using a constraint degree similar to the total MB-VIOP ≥ 0.5, and one for the reduced model using a constraint degree similar to the total MB-VIOP ≥ 1

ISSN: 1471-2105