Variablea N Estb Average Estc Range Estd Average SEe
Binary:        
HRK3 10 -0.1264 0.0644 0.0034
HRK2 9 -0.0789 0.0284 0.0036
HRK1 8 -0.1023 0.0422 0.0042
P1 10 0.0783 0.0300 0.0030
P2 10 -0.0719 0.0281 0.0030
P3 10 -0.0798 0.0255 0.0030
H3 8 0.0492 0.0445 0.0049
Sulfur3 9 0.0825 0.1389 0.0112
Sulfur2 8 0.0350f 0.0316 0.0049
Continuous:        
Relative Fragment Size 10 -0.6955 0.1748 0.0241
Relative Fragment Mass 10 0.8009 0.1693 0.0250
aVariables significant at p-value < 0.00005 in seven or more data sets and minimum average estimates ≥ 0.05 log10 (ion intensity); Binary explanatory variable nomenclature: X#: X denotes the residue (individual letter) or residue group (word or group of letters corresponding to the residues), and # denotes the position of the residue or group relative to the production ion fragmentation site (e.g. C2 denotes C located at the second to last position from the fragmentation terminus); HRK: group of residues that have positive charge and are large and hydrophilic (H, R, and K); P: denotes residue P that has neutral charge, is hydrophilic, imino and tiny; N: denotes residue N that has neutral charge, is hydrophilic, amide and small; H: denotes residue H that has a positive charge, is hydrophilic and large; Sulfur: group of residues that have sulfur (C, and M); HRK: group of residues that have positive charge, and are large and hydrophilic; Relative Fragment Size: ratio between the fragment ion and peptide ion length; Relative Fragment Mass: ratio between the fragment ion and peptide ion mass.
bNumber of training data sets significant at p-value < 0.00005.
c
Average of the log10 (intensity) estimates, adjusted for all other variables in the model, across the number of data sets in which the variable appeared. For the binary variables estimate is the difference in log10 (ion intensity) between presence and absence of the residue or residue group at a specific location. For the continuous variables estimate is the change in ion intensity per unity change in the continuous variable.
dRange of the log10 (intensity) estimates across the number of data sets in which the variable appeared.
e
Average standard error of the log10 (intensity) estimates across the number of data sets in which the variable appeared.
f
Borderline chemically significant.
Table 3: Average estimates and standard errors of the binary and continuous explanatory variables significantly associated with fragment ion intensity and number of data sets (out of 10 sets) that detected the variable.