Table 4 Influence of factors specifying the pairwise peak-list comparison on partial areas under the ROC curve for intensity PMF and MS/MS data. For each of the 2688 pairwise peak-list comparison approaches, sensitivity-PAUC (sensitivity given FP-rate ∈ [0, 0.1]) and specificity-PAUC (specificity given sensitivity ∈ [0.9, 1]) (Figure 2) were determined. A partitioning of sums of squares was performed analogously to analysis of variance. Column names: Factors – identification of factors; df – degrees of freedom (DF, number of factor levels - 1); %SSQ – relative sum of squares (%SSQ = SSQ/∑SSQ); %MSQ – relative mean sum of squares (%MSQ = MSQ/∑MSQ), where %MSQ = MSQ/∑MSQ. %MSQ measures the importance of a specific factor for the size of sensitivity-PAUC and specificity-PAUC. × denotes interactions between factors. measure – distance measure, noncross – non crossing matching, length – alignment length, θ – weight of non-matching peaks, trans – peak intensity transformation, residual – unexplained %SSQ or %MSQ, total – column sum of %SSQ, df, %MSQ.

From: Transformation and other factors of the peptide mass spectrometry pairwise peak-list comparison process