Dear David,
to generate my library I have done a couple of steps:
1) modified the file utils.py from seq2ms to contain my modifications:
*# C H N O S P# mods = {"ox" : [0,0,0,1,0,0], ## changed this line to
enable sulfatase modification prediction se=serine;
al=formylglycine-aldehyd; do=formylglycine-diol; ds=diol-sulfate;
ss=serine-sulfatemods = {"ox" : [0,0,0,1,0,0], "se" : [0,0,1,0,-1,0], "al"
: [0,-2,1,0,-1,0], "do" : [0,0,2,0,-1,0], "ds" : [0,-1,5,0,0,0], "ss" :
[0,0,4,0,0,0], "ph" : [0,1,0,3,0,1], "cam" : [2,3,1,1,0,0] ,
"ac": [2,2,0,1,0,0], "me": [1,2,0,0,0,0], "hy": [0,0,0,1,0,0], "gly":
[4,6,2,2,0,0], "bi" : [10,14,2,2,1,0], "cr": [4,4,0,1,0,0], "di":
[2,4,0,0,0,0], "ma": [3,2,0,3,0,0], "ni": [0,-1,1,2,0,0], "bu" :
[4,6,0,1,0,0], "fo": [1,0,0,1,0,0], "glu": [5,6,0,3,0,0], "hyb":
[4,6,0,2,0,0], "pr": [3,4,0,1,0,0], "su" : [4,4,0,3,0,0], "tr":
[3,6,0,0,0,0], "ci": [0,-1,-1,1,0,0]}*
2) I run the prediction by executing 'predict.py' with the a modified
input.tsv and (I think) the pretrained_model. The input file looked like
the following but contained more peptides:
*Sequence Charge Mass Modified sequence Modification
ProteinDILTPELDNLAQNGSIFTSAYVAHPFCGPSR 2 3332.613578
_DILTPELDNLAQNGSIFTSAYVAHPFCGPSR_ Unmodified
DILTPELDNLAQNGSIFTSAYVAHPFCGPSR 3 3332.613576
_DILTPELDNLAQNGSIFTSAYVAHPFCGPSR_ Unmodified
DILTPELDNLAQNGSIFTSAYVAHPFCGPSR 4 3332.613576
_DILTPELDNLAQNGSIFTSAYVAHPFCGPSR_ Unmodified
DILTPELDNLAQNGSIFTSAYVAHPFCGPSR 2 3316.636422
_DILTPELDNLAQNGSIFTSAYVAHPFC(se)GPSR_ Serine (C)
DILTPELDNLAQNGSIFTSAYVAHPFCGPSR 3 3316.636422
_DILTPELDNLAQNGSIFTSAYVAHPFC(se)GPSR_ Serine (C)
DILTPELDNLAQNGSIFTSAYVAHPFCGPSR 4 3316.63642
_DILTPELDNLAQNGSIFTSAYVAHPFC(se)GPSR_ Serine (C)
DILTPELDNLAQNGSIFTSAYVAHPFCGPSR 2 3314.620772
_DILTPELDNLAQNGSIFTSAYVAHPFC(al)GPSR_ FGAldehyd (C)
DILTPELDNLAQNGSIFTSAYVAHPFCGPSR 3 3314.62077
_DILTPELDNLAQNGSIFTSAYVAHPFC(al)GPSR_ FGAldehyd (C)
DILTPELDNLAQNGSIFTSAYVAHPFCGPSR 4 3314.620772
_DILTPELDNLAQNGSIFTSAYVAHPFC(al)GPSR_ FGAldehyd (C)
DILTPELDNLAQNGSIFTSAYVAHPFCGPSR 2 3332.631336
_DILTPELDNLAQNGSIFTSAYVAHPFC(do)GPSR_ FGDiol (C)
DILTPELDNLAQNGSIFTSAYVAHPFCGPSR 3 3332.631336
_DILTPELDNLAQNGSIFTSAYVAHPFC(do)GPSR_ FGDiol (C)
DILTPELDNLAQNGSIFTSAYVAHPFCGPSR 4 3332.631336
_DILTPELDNLAQNGSIFTSAYVAHPFC(do)GPSR_ FGDiol (C)
DILTPELDNLAQNGSIFTSAYVAHPFCGPSR 2 3411.580326
_DILTPELDNLAQNGSIFTSAYVAHPFC(ds)GPSR_ DiolSulfat (C)
DILTPELDNLAQNGSIFTSAYVAHPFCGPSR 3 3411.580326
_DILTPELDNLAQNGSIFTSAYVAHPFC(ds)GPSR_ DiolSulfat (C)
DILTPELDNLAQNGSIFTSAYVAHPFCGPSR 4 3411.580324
_DILTPELDNLAQNGSIFTSAYVAHPFC(ds)GPSR_ DiolSulfat (C)
DILTPELDNLAQNGSIFTSAYVAHPFCGPSR 2 3396.593236
_DILTPELDNLAQNGSIFTSAYVAHPFC(ss)GPSR_ SerinSulfat (C)
DILTPELDNLAQNGSIFTSAYVAHPFCGPSR 3 3396.593235
_DILTPELDNLAQNGSIFTSAYVAHPFC(ss)GPSR_ Unmodified
DILTPELDNLAQNGSIFTSAYVAHPFCGPSR 4 3396.593236
_DILTPELDNLAQNGSIFTSAYVAHPFC(ss)GPSR_ Unmodified *
3) I imported the generated .msp file into spectraST format using the
command line and the following options:
*C:\TPP\bin\spectrast.exe -cNPredicted_Sec2MS.splib -MSec2MS.usermods
Predicted.msp*
4) I tried to visualise this library by Lib2HTML from the within Petunia
For the files I will send you a link via the ISB contact form to our
university's nextcloud where I have uploaded them.
Best regards,
Juergen
Juergen Bartel schrieb am Montag, 22. Januar 2024 um 18:22:49 UTC+1:
> Dear all,
>
> I have recently generated an in-silico spectral library for several
> peptides which may contains different modifications on the same cystein
> (cystein converted to a serin, formylglycin-aldehyde, ...). For this
> prediction I used Seq2MS (
> https://pubs.acs.org/doi/10.1021/acs.jproteome.3c00180) and obtained a
> library in msp format.
> In this file the header indicates modifications for examle in the
> following way:
>
> Name: DILTPELDNLAQNGSIFTSAYVAHPFCGPSR/2_1(26,C,FGAldehyd)
> Comment: Charge=2 Parent=1657.310386 Mods=1(26,C,FGAldehyd) Protein=nan
>
> If I correctly specify the modifications in a .usermods file(*) and import
> it to spectraST, the resulting sptxt files contain header such as:
>
> Name: DILTPELDNLAQNGSIFTSAYVAHPFC[85]GPSR/2
> LibID: 6
> MW: 3316.6353
> PrecursorMZ: 1658.3177
> Status: Normal
> FullName: X.DILTPELDNLAQNGSIFTSAYVAHPFC[85]GPSR.X/2 (CID)
> Comment: AvePrecursorMz=1659.3531 BinaryFileOffset=19516 Charge=2
> FracUnassigned=0.89,4/5;0.83,16/20;0.56,35/60 Mods=1(26,C,FGAldehyd) NAA=31
> NISTProtein=nan NMC=0 NTT=1 Parent=1657.310386 Prob=1.0000 Protein=1/nan
>
> However, when I want to check the spectra in this file in HTML format
> using Lib2HTML, the table contains the complete peptide only for the
> unmodified form and has an identical truncated peptide for all other
> modified forms:
>
> [image: Unbenannt.PNG]
>
> Ths does not only affect the truncated amino acids but also results in
> mis-alignment of the y-ion-series (i.e. due to the modified cystein, y1 is
> annotated as 122.027 while in this case it should be y5 with a much larger
> m/z). Again, the annotation in sptxt seems good.
>
> Does anyone has an idea how to solve this and/or how I could watch the
> spectra alternatively?
>
> Best,
> Juergen
>
>
> (*) part of the usermods file:
> C[se]|-15.977156|Serin
> C[al]|-17.992806|FGAldehyd
> ...
>
--
You received this message because you are subscribed to the Google Groups
"spctools-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to spctools-discuss+unsubscr...@googlegroups.com.
To view this discussion on the web visit
https://groups.google.com/d/msgid/spctools-discuss/dd0ac502-dde9-495c-ac48-58340707b5b9n%40googlegroups.com.
