3.3. False positives
The majority of FP results were due to different cyclic peptides
similar in structure to ‘true’ compounds related to test spectra. It
can be seen for the case of searches among literature spectra
resulted in TPR of 75% (see Table 3). There were six FPs in positive
control and four FPs in negative one recorded in test searches for
this sublibrary. In the first group, the difference in structure
between the test compound as TP and the falsely identified ones
is the only amino acid, i.e. it is the minor difference (Table 5). In
the second case, compared structures are the same, with different
precursor ions, or are differed by the substructure ranging from
hydroxyl to the three amino acid residue (Table 5).
Linear peptides sometimes occupied the 1st position in hit-lists.
All corresponding MFs were low and under MFcut-off thresholds for
positive results. Correspondingly, there were no linear compounds
among total of 23 false positive results obtained at positive and
negative control during the validation of the entire library. Thus
false identification of an individual cyclic peptide implies true
group recognition of cyclic peptides as the class of natural compounds
and even its subclass (e.g. microcystins, see Table 5).
3.3. False positives
The majority of FP results were due to different cyclic peptides
similar in structure to ‘true’ compounds related to test spectra. It
can be seen for the case of searches among literature spectra
resulted in TPR of 75% (see Table 3). There were six FPs in positive
control and four FPs in negative one recorded in test searches for
this sublibrary. In the first group, the difference in structure
between the test compound as TP and the falsely identified ones
is the only amino acid, i.e. it is the minor difference (Table 5). In
the second case, compared structures are the same, with different
precursor ions, or are differed by the substructure ranging from
hydroxyl to the three amino acid residue (Table 5).
Linear peptides sometimes occupied the 1st position in hit-lists.
All corresponding MFs were low and under MFcut-off thresholds for
positive results. Correspondingly, there were no linear compounds
among total of 23 false positive results obtained at positive and
negative control during the validation of the entire library. Thus
false identification of an individual cyclic peptide implies true
group recognition of cyclic peptides as the class of natural compounds
and even its subclass (e.g. microcystins, see Table 5).
การแปล กรุณารอสักครู่..
