A calibrated and validated discriminating rule built on the combination of the data obtained from the two MALDI-FTICR methods resulted in a sensitivity of 89% and a specificity of 100% with an AUC of 0.989. These results corroborate classification

numbers from our previous MALDI-TOF studies [19] and [32]. The t-test analysis performed on the peptides with absolute discriminant weights higher than 0.1 resulted in the identification of 34 peptides Alectinib mouse that (i.e. p-value lower than 0.001) differentiate between case and control groups (see Table 3). The high precision and accuracy of the mass measurements allowed the identification of 26 of these peptides either by comparison with previously reported peptides or by accurate mass measurement of mass differences in the spectra (see Section 2). Application of the latter approach resulted in the identification of peptides generated through proteolysis of the same protein. In fact, starting from a previously identified peak (i.e. peptide) it was found that accurate measurement of the difference between that specific m/z-value and the m/z-value GDC-0449 in vivo of a new peak matched to a similar peptide with either one amino acid more or less at the C- or the N-terminus, corresponding to the “overall” protein sequence. Thus,

up to 8 new peptides could be identified starting from the fragment peptide K.SLEDKTERELLESYIDGR of thrombin light chain (UniProt P00734) (see Table 2). Nevertheless, the presence of isobaric peptides cannot be excluded and MS/MS experiments are required to further validate the identifications. In conclusion, using the two identification approaches described above,

we are now able to further expand the total number of identified peptides, especially at higher m/z-values. Other MALDI-profiling methods that so far have been used for the characterization of human serum peptides were not suitable for the identification of high molecular weight peptides or proteins, Dapagliflozin because these lacked sensitivity and resolving power [28] and [29]. As a final remark, it should be noted that at this stage the peptidome profiles were not evaluated for the m/z-range from 9000 to 10,000. Here, both the high density of peaks and the relatively lower resolving power do not permit binning of the data points. The most abundant peaks present in this range were identified as apolipoprotein-CIII isoforms [26] and these data will be evaluated in a separate study using a different quantification method. In this study, we have shown that high quality human serum peptide and protein profiles can be generated using a standardized and robust protocol for the sample preparation and ultrahigh resolution 15 T MALDI-FTICR MS for the mass measurements.