PepNovo doesn't apply fixed modifications correctly

[airizzo]

In PepNovo results generated with DeNovo GUI (latest version), fixed modifications are applied like variable modifications and only show up some of the time, despite being specified as fixed. For example, in the results below (from the PepNovo .out file), which was run with parameters specifying the fixed modifications n-terminal dimethylation and carbamidomethylated cysteine, many spectra show up with sequences containing neither. With these specifications, every single result should have the dimethylation and every cysteine should be modified.

Modifications worked as expected for a Novor analysis run in the same session (same spectra and settings files). I'm wondering if there is a bug in how the GUI passes on modification info to PepNovo.

Non-application of fixed modifications:

>> 0 9747 File: "D:\Orbitrap data\2015\10-Oct2015\16Oct2015\DvH1-IPTL-18Oto16O-Oct16-6.raw"; SpectrumID: "2878"; scans: "9747" (SQS 0.996)
#Index  RnkScr  PnvScr  N-Gap   C-Gap   [M+H]   Charge  Sequence
0   -1.215  68.328  275.094 0.000   1317.678    2   TSHTCVPLK
1   -1.710  68.424  293.198 0.000   1317.678    2   SSHELNPLK
2   -1.929  71.508  188.052 0.000   1317.678    2   STSHTCVPLK
3   -1.947  75.181  293.198 0.000   1317.678    2   SSHEAGVPLK
4   -2.058  71.230  248.153 0.000   1317.678    2   QSHTCVPLK
5   -2.204  68.000  188.052 0.000   1317.678    2   CVELGDVPLK
6   -2.414  68.999  0.000   0.000   1317.678    2   STSTAELDRPLK
7   -2.420  20.453  345.160 0.000   1314.683    2   VFTLTEQM
8   -2.470  22.932  345.160 0.000   1314.683    2   VFTTLEQM
9   -2.513  74.233  275.094 0.000   1317.678    2   TAELGDVPLK

Inconsistent application of fixed modifications:

>> 0 9750 File: "D:\Orbitrap data\2015\10-Oct2015\16Oct2015\DvH1-IPTL-18Oto16O-Oct16-6.raw"; SpectrumID: "2880"; scans: "9750" (SQS 0.676)
#Index  RnkScr  PnvScr  N-Gap   C-Gap   [M+H]   Charge  Sequence
0   0.249   -21.184 130.600 0.000   1320.584    3   AMAGSFYCQR
1   0.078   -21.314 130.600 0.000   1320.584    3   AMASGFYCQR
2   -0.068  -24.494 130.600 0.000   1320.584    3   AMAGSFYCKR
3   -2.339  23.939  0.000   0.000   1323.580    3   AACTVADATGSGSR
4   -2.356  33.587  0.000   0.000   1323.580    3   Q^+28EEHHATGGSSR
5   -2.510  33.238  0.000   0.000   1323.580    3   Q^+28EEHHATGSGSR
6   -2.543  22.780  0.000   0.000   1323.580    3   N^+28CTVADATGVCK
7   -2.547  17.589  0.000   0.000   1323.580    3   ADVEHHACAASR
8   -2.556  17.626  0.000   0.000   1323.580    3   AVDEHHACAASR
9   -2.595  27.553  0.000   0.000   1323.580    3   Y^+28DVPADACTQR

[airizzo]

I have a few more insights into this bug. Here is the PepNovo_PTMs.txt file, which was presumably rewritten by DeNovo GUI:

#AA  offset      type    locations  symbol  PTM name
N_TERM  28.0313 OPTIONAL N_TERM ^+28 DIMETHYLATION OF PEPTIDE N-TERM
C   57.021464 FIXED ALL C+57 CARBAMIDOMETHYLATION OF C
M   15.994915 OPTIONAL ALL M+16 OXIDATION OF M

The problem here seems to be that the n-term modification here is listed as "OPTIONAL", even though it is specified as fixed in the parameters file. I've attached the parameters file here so you can verify this. IPTLsettings12-14.zip

It looks like rather than passing the fixed modification all the way down, De Novo GUI is randomly applying modification labels at the top level, and possibly not changing masses accordingly. This screen shot displays the results view in the GUI for the first spectrum posted in OP. Although in the PepNovo results file, none of the sequences are listed as having n-terminal dimethylation, according to DeNovo GUI one, but not all of the sequences are.

Hope there's an easy fix for this.

[hbarsnes]

Hi,

I'm afraid the modification support in PepNovo itself is not as complete as one could hope for. Fixed terminal PTMs are sadly not supported. Thus the only way we can support such modifications in DeNovoGUI is to rather add them as variable terminal modifications. Not optimal, but at least it allows you to search for them.

As to the fixed non-terminal modifications, such as carbamidomethylated cysteine, these are not annotated in the PepNovo output file. It is rather assumed that one remembers the search parameters and that carbamidomethylated cysteine is re-added when displaying the data or calculating the peptide mass. This is what we do in DeNovoGUI.

Note also that fixed non-terminal modifications are hidden in DeNovoGUI by default. To show them, go to the View menu and make sure that "Fixed Modifications" is selected. The carbamidomethylated cysteines will then be annotated in the peptide sequences in the table.

I hope this answers your questions?

Best regards, Harald

[WaldbauerLab]

Sorry to hear that. Thanks for the clarification.