Table 2.

Efficiency of the prediction of Pro/Hyp location in CWP amino acid sequences according to the proposed rules.

Accession number	Number of analyzed Pro/Hyp positions	Number of unexpected location of Pro/Hyp residues	Percentage of mis-predictions
At4g38770	868	44	5.1
At1g09750	343	104	30.3
At1g31580	439	154	35.1
At3g08030	502	225	44.8
At2g10940	602	281	46.7

Five CWP sequences were analyzed by LC-MS/MS (Supplementary Table S1) and the percentage of mis-predictions was calculated (Supplementary Table S2).