-
Notifications
You must be signed in to change notification settings - Fork 5
3. Output
global.profile.tsv reports the most prominent features from PTM-Shepherd analysis of mass shifts observed from FDR-filtered open search results. Each row corresponds to a different detected mass shift, thus not all PSMs will be represented in this table. Please note that mass shifts are annotated based on UniMod mapping, thus they are not definitive chemical identities and should be used as a starting point along with localization and amino acid enrichment information. Unless otherwise indicated, values are summed from all datasets in the analysis. Column contents are listed below.
peak_apex apex of the detected delta mass peak (in Da)
peak_lower lower bound of the detected peak (Da), determined by precursor tolerance or the detection of an adjacent peak
peak_upper upper bound of the detected peak (Da), determined by precursor tolerance or the detection of an adjacent peak
PSMs the number of PSMs contained within the peak boundary (bin), reported for each dataset if multiple datasets are used as input
peak_signal relative measure of peak prominence/quality. In noisy regions of the delta mass histogram, values are penalized
percent_also_in_unmodified the percentage of PSMs in this mass bin with a corresponding PSM in the unmodified bin
mapped_mass_1 primary modification annotation derived from Unimod, all isobaric modifications listed and separated by “/”
mapped_mass_2 if the delta mass peak is a combination of two masses, a second modification annotation is listed here. As with mapped_mass_1, all isobaric modifications are listed and separated by “/”
similarity MS/MS spectral similarity of modified peptides compared to their unmodified counterparts. When multiple modified-unmodified comparisons are done for a single peptide, these cosine similarity scores are averaged for the peptide. The peptide scores are then averaged across all peptides in the mass shift bin. These comparisons are only done for peptides of the same charge state.
rt_shift retention time shift comparing modified peptides to their unmodified counterparts. When multiple modified-unmodified comparisons are done for a single peptide, the retention time shifts are averaged for the peptide. The peptide shifts are then averaged across all peptides in the mass shift bin. Individual comparisons are only done for peptides in the same LC-MS run. Units are usually seconds but can vary by instrument type
int_log2fc log2 fold-change of average intensity for matched shifted/unshifted peptides, computed as described above. Peptides affect by sample preparation artifacts tend to be lower abundance than their unshifted counterparts, thus this value will be low in these cases
localized_PSMs number of PSMs for this delta mass that showed at least one additional matched ion when the mass shift is placed on a residue
n-term_localization_rate percentage of PSMs with an uninterrupted string of localized residues from the N-terminus. This is calculated differently from other enrichment scores due to the difference in assumptions underlying N-terminal and residue-specific localization, so these values cannot be directly compared to the amino acid enrichment scores.
AA1 amino acid/residue most enriched (most likely to harbor the mass shift) compared to other residues
AA1_enrichment_score equivalent to the odds the delta mass is localized to AA1 compared to other residues
AA1_psm_count weighted number of PSMs where the mass shift localized to AA1. Shifts localizing to multiple residues are divided by the number of localized residues in the spectra, so this is an estimated number of PSMs localized to a particular residue
(same enrichment_score, and psm_count columns for AA2 and AA3 if multiple amino acids are likely to harbor the mass shift)
[experiment]_PSMs number of PSMs with a mass shift in this bin
[experiment]_percent_PSMs number of PSMs from the previous column as a percentage of total PSMs
[experiment]_peptides number of unique peptide sequences with a mass shift in this bin
[experiment]_percent_also_in_unmodified percentage of peptide sequences with a mass shift in this bin that are also found in the zero mass shift bin
global.diagmine.tsv is a mass shift-centric table that contains the diagnostic features identified for every mass shift. Please note that only mass shifts with diagnostic features detected are reported in the table. Contents of each column are listed below.
peak_apex This field contains the apex of the detected MS1 peak (Da) present in the global.profile.tsv file from PTM-Shepherd.
mod_annotation This field contains the mass shift annotations present in the global.profile.tsv file from PTM-Shepherd. When a mass shift is found to be the combination of two mass shifts, the “Potential Modification 1” and “Potential Modification 2” columns are merged with a semicolon.
type This field can take one of several values. “diagnostic” refers to diagnostic ions, the ions that can be located directly in the spectrum. “peptide” refers to peptide remainder masses, mass shifts that indicate an ion’s presence at a particular distance from an unshifted peptide. Six other values are possible based on parameter setting, each corresponding to one of the major ion series.
mass This field contains the mass of the diagnostic feature. Peptide and fragment remainder masses will have the mass shift away from the theoretical ion. Diagnostic ions will have the m/z of the observed ion, so a non-neutral mass.
delta_mod_mass This field contains the mass that was lost from the original mass shift to arrive at the remainder mass. (Note: only present for peptide and fragment remainder masses.)
remainder_propensity This field contains the average percentage of ions from a particular series that are shifted. For example, a peptide capable of producing 10 b-ions with 2 ions identified ions shifted by the remainder mass and 2 identified ions unshifted would have a propensity of 50%. The propensity score for every representative PSM within a mass shift bin is averaged. (Note: only present for fragment remainder masses.)
percent_mod This field contains the percentage of representative mass shifted PSMs that contain the ion at any intensity.
percent_unmod This field contains the percentage of representative unshifted PSMs that contain the ion at any intensity.
avg_intensity_mod This field contains the average intensity of the ion among representative mass shifted PSMs where the ion is present. To calculate the average across all representative mass shifted spectra, calculate (avg_intensity_mod * percent_mod / 100). Because multiple ions can be matched for fragment remainder ions, this contains the average of the summed intensity of matched ions for each representative PSM.
avg_intensity_unmod This field contains the average intensity of the ions among representative unshifted PSMs where the ion is present. To calculate the average across all representative mass shifted spectra, calculate (avg_intensity_mod * percent_mod / 100). Because multiple ions can be matched for fragment remainder ions, this contains the average of the summed intensity of matched ions for each representative PSM.
intensity_fold_change This field contains the fold change in intensity when comparing the modified to unmodified peptides. This uses intensity across all spectra and can be calculated via (avg_intensity_mod * percent_mod) / (avg_intensity_unmod * percent_unmod).
auc This column contains the AUC-ROC statistic for the intensity-based classification of this ion. It is calculated from the U statistic from the Mann-Whitney U Test. This statistic adjusts the two groups such that they are assumed to be of equal size.
global.modsummary.tsv is a modification-centric table generated from PTM-Shepherd summarization of mass shifts observed in open search workflows. Please note that mass shifts are annotated based on UniMod mapping, thus they are not definitive chemical identities and should be used as a starting point along with localization and amino acid enrichment information. Contents of each column are listed below.
Modification Name/annotation of the modification (as found in the global.profile.tsv file)
Theoretical Mass Shift The theoretical mass (in Da) of the modification from Unimod if annotated, or the peak apex of an unannotated modification
[experiment]_PSMs Number of PSMs with the modification, including any row from the global.profile.tsv file where the modification appears (e.g., a 'Methylation' entry in the will include PSMs corresponding to both 'Methylation' and 'Methylation + First isotopic peak')
[experiment]_percent_PSMs The number of PSMs from the previous column as a percentage of the total PSMs