Hello Piero,

I have not looked at this, but I suspect it could be dataset related.
Localizations at position 1 could indicate peptide extensions, and I am not
sure why positions 3 and 4 are so populated.  Have you tried looking at
this while controlling for peptides of various lengths? If you can post
your data in a share I could take a closer look.

Thanks,
-David

On Sun, Jul 8, 2018 at 11:51 PM, Piero Giansanti <[email protected]>
wrote:

> Hello David,
>
> I am back with an additional question :)
> I have noticed that in my data set the vast majority of the mass shifts
> (outside the +/-0.02 Da window) are localized withing the fist 4 residues
> (from the n-term), except for position #2.
> This is for a localization probability > 0.75, but it does not change if I
> use > 0.9, except for # 3 and 4 that tend to be slightly less frequent that
> position #1.
> Have you noticed the same?
>
> Thanks
> -Piero
>
>
> <https://lh3.googleusercontent.com/-pd1QudrhPDQ/W0MEr8yj6ZI/AAAAAAAAALQ/aPqqd0KQjBcjqI5-h7PAFnm5TrhL4dQXwCLcBGAs/s1600/loc.jpg>
>
>
>
>
> On Monday, June 11, 2018 at 8:57:53 PM UTC+2, David Shteynberg wrote:
>>
>> Hello Piero.
>>
>> The reason mass shift localizations are performed on all amino acids is
>> to avoid any assumptions on the part of PTMProphet.  The goal of PTMProphet
>> in mass shift localization mode is to simply localize the mass shift
>> without assuming anything else about the modification.  This can help the
>> user discover rare modifications.  For instance, phosphorylation on D is
>> listed as a rare modification in unimod.
>>
>> Cheers,
>>
>> -David
>>
>> On Mon, Jun 11, 2018 at 9:40 AM, Piero Giansanti <[email protected]>
>> wrote:
>>
>>> That worked out, eventually :)
>>>
>>> I have an additional question though, which comes from the fact that the
>>> majority of the known mass shift in my data set has not been fully
>>> localized on a specific residue, even when spectra look OK-ish.
>>> My background is totally not bioinformatics/statistics, so this could be
>>> a silly question :) ...
>>>
>>> Anyway, is there a reason why the localization of the mass shift is
>>> performed differently compared the the localization of a known modification?
>>> What I mean is that (as far as I understood), when one wants to perform
>>> localization on (phospho.STY)PEPTIDE, PTMprophet will simply return T4 as
>>> the phosphorylated residues (it will not even try to look at D6),
>>> but when trying on (+79.966Da)PEPTIDE, PTMprophet will not focus on just
>>> T and D, but rather 'look' at the entire sequence.
>>>
>>> Wouldn't this 'focusing' speed up the time need for the localization,
>>> and boost the localization rate?
>>>
>>> Thanks
>>> -Piero
>>>
>>>
>>>
>>>
>>> On Tuesday, May 22, 2018 at 6:43:33 PM UTC+2, David Shteynberg wrote:
>>>>
>>>> If you are able to build the TPP the code has been checked into the SVN
>>>> repository.  You can find the code here:
>>>>
>>>> https://sourceforge.net/p/sashimi/code/HEAD/tree/trunk/trans
>>>> _proteomic_pipeline/
>>>>
>>>> If not I can send you a precompiled executable.  Which linux flavor are
>>>> you using?
>>>>
>>>> -David
>>>>
>>>> On Tue, May 22, 2018 at 5:40 AM, Piero Giansanti <[email protected]>
>>>> wrote:
>>>>
>>>>> Hello David,
>>>>>
>>>>> Just tried with TPP 5.1.0 (linux), but same problem as you can see
>>>>> below.
>>>>> Is there a way to get hands on the version you have used? :)
>>>>>
>>>>> <spectrum_query start_scan="3214" assumed_charge="2"
>>>>> spectrum="01928_GA1_P020123_S00_R01_R1.03214.03214.2" end_scan="3214"
>>>>> index="6" precursor_neutral_mass="1247.5479"
>>>>> retention_time_sec="1120.817">
>>>>> <search_result>
>>>>> <search_hit peptide="GGMNGASPPTSQK" massdiff="0.9869"
>>>>> calc_neutral_pep_mass="1246.5610" peptide_next_aa="D"
>>>>> num_missed_cleavages="0" num_tol_term="2" num_tot_proteins="1"
>>>>> tot_num_ions="24" hit_rank="1" num_matched_ions="10"
>>>>> protein="sp|Q62234|MYOM1_MOUSE" peptide_prev_aa="R" is_rejected="0"
>>>>> protein_descr="Myomesin-1 OS=Mus musculus OX=10090 GN=Myom1 PE=1 SV=2">
>>>>> <modification_info modified_peptide="G[58]GM[147]NGASPPTSQK">
>>>>> <mod_aminoacid_mass position="1" mass="58.00836"/>
>>>>> <mod_aminoacid_mass position="3" mass="147.0354"/>
>>>>> </modification_info>
>>>>> <search_score name="hyperscore" value="21.034"/>
>>>>> <search_score name="nextscore" value="18.157"/>
>>>>> <search_score name="expect" value="1.601e-03"/>
>>>>> <analysis_result analysis="peptideprophet">
>>>>> <peptideprophet_result probability="0.9995"
>>>>> all_ntt_prob="(0.9840,0.9981,0.9995)">
>>>>> <search_score_summary>
>>>>> <parameter name="fval" value="5.8278"/>
>>>>> <parameter name="ntt" value="2"/>
>>>>> <parameter name="nmc" value="0"/>
>>>>> <parameter name="massd" value="0.987"/>
>>>>> </search_score_summary>
>>>>> </peptideprophet_result>
>>>>> </analysis_result>
>>>>> <analysis_result analysis="ptmprophet">
>>>>> <ptmprophet_result prior="0.0769231" 
>>>>> ptm="PTMProphet_ABCDEFGHIJKLMNOPQRSTUVXWYZ0.9869"
>>>>> ptm_peptide="G(0.000)G(0.000)M(0.000)N(0.000)G(0.000)A(0.000
>>>>> )S(0.000)P(0.000)P(0.000)T(0.000)S(0.000)Q(0.000)K(0.000)">
>>>>> <mod_aminoacid_probability position="1" probability="0.000"/>
>>>>> <mod_aminoacid_probability position="2" probability="0.000"/>
>>>>> <mod_aminoacid_probability position="3" probability="0.000"/>
>>>>> <mod_aminoacid_probability position="4" probability="0.000"/>
>>>>> <mod_aminoacid_probability position="5" probability="0.000"/>
>>>>> <mod_aminoacid_probability position="6" probability="0.000"/>
>>>>> <mod_aminoacid_probability position="7" probability="0.000"/>
>>>>> <mod_aminoacid_probability position="8" probability="0.000"/>
>>>>> <mod_aminoacid_probability position="9" probability="0.000"/>
>>>>> <mod_aminoacid_probability position="10" probability="0.000"/>
>>>>> <mod_aminoacid_probability position="11" probability="0.000"/>
>>>>> <mod_aminoacid_probability position="12" probability="0.000"/>
>>>>> <mod_aminoacid_probability position="13" probability="0.000"/>
>>>>> </ptmprophet_result>
>>>>> </analysis_result>
>>>>> </search_hit>
>>>>> </search_result>
>>>>> </spectrum_query>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> On Friday, May 18, 2018 at 6:58:36 PM UTC+2, David Shteynberg wrote:
>>>>>>
>>>>>> Hello Piero,
>>>>>>
>>>>>> It appears that the version of PTMProphet in Philosopher is a bit out
>>>>>> of date.  The latest TPP version is 5.1.0, and PTMProphet has been 
>>>>>> updated
>>>>>> with bugfixes and new features in the meantime. I have run my latest
>>>>>> PTMProphet code (targeted for the 5.2.0 version of TPP) on your data.  
>>>>>> For
>>>>>> the PSMs matching this peptide I cannot replicate the issue you've
>>>>>> reported.  Instead, PTMProphet seems to be positioning the mass
>>>>>> modification on the Asparagine in your peptide as you should see in the
>>>>>> following result file:
>>>>>>
>>>>>> https://www.dropbox.com/s/z3xf379ulljr3rr/01928_GA1_P020123_
>>>>>> S00_R01_R1.PTMPro.pep.xml?dl=0
>>>>>>
>>>>>> Thanks for your interest in PTMProphet.
>>>>>>
>>>>>> Cheers!
>>>>>>
>>>>>> -David
>>>>>>
>>>>>> On Fri, May 18, 2018 at 8:40 AM, Piero Giansanti <
>>>>>> [email protected]> wrote:
>>>>>>
>>>>>>> Hello David,
>>>>>>>
>>>>>>> I am using PTMprophet via philosopher, and from the xml output it
>>>>>>> seems it is this version
>>>>>>> TPP v5.0.1 Post-Typhoon dev, Build 201706282113-exported
>>>>>>> (Linux-x86_64)
>>>>>>>
>>>>>>> Regarding the settings, I am using --massdiffmode --mztol 0.01. The
>>>>>>> mass tolerance is not exactly the one used for the database search (20
>>>>>>> ppm), but it should be somehow close to it.
>>>>>>>
>>>>>>> Here https://ufile.io/wiqqb the mzML, pepXML and the output I
>>>>>>> generated.
>>>>>>>
>>>>>>> Thanks
>>>>>>>
>>>>>>>
>>>>>>> On Friday, May 18, 2018 at 4:21:27 PM UTC+2, David Shteynberg wrote:
>>>>>>>>
>>>>>>>> Hello Piero,
>>>>>>>>
>>>>>>>> Thank you for trying the software.  This is indeed an unusual
>>>>>>>> result.  Can you tell me which version of PTMProphet you are using and 
>>>>>>>> how
>>>>>>>> you are running the tool?  Is it possible for you to post your pepXML 
>>>>>>>> file
>>>>>>>> and data file online for me to download and test this particular ID?
>>>>>>>>
>>>>>>>> Cheers,
>>>>>>>> -David
>>>>>>>>
>>>>>>>> On Fri, May 18, 2018 at 1:53 AM, Piero Giansanti <
>>>>>>>> [email protected]> wrote:
>>>>>>>>
>>>>>>>>> Hi,
>>>>>>>>>
>>>>>>>>> I am using PTMprophet to localize the set of modifications (i.e.
>>>>>>>>> mass shifts) identified by running an open search with MSfragger, but 
>>>>>>>>> I
>>>>>>>>> have some doubt regarding the interpretation of the output.
>>>>>>>>> In particular, MSfragger identified this peptide
>>>>>>>>> *GG(ox)MNGASPPTSQK*
>>>>>>>>> with a mass shift of 0.9869 Da, which match to the following
>>>>>>>>> modification according to Unimod
>>>>>>>>> 0.9840:Deamidated (Deamidation), 0.9840:Asn->Asp (Asn->Asp
>>>>>>>>> substitution), 0.9840:Gln->Glu (Gln->Glu substitution).
>>>>>>>>>
>>>>>>>>> Of these 3 only the first 2 are possible (there is no Q in the
>>>>>>>>> sequence, so there cannot be any Gln->Glu substitution).
>>>>>>>>>
>>>>>>>>> Now, the output of PTMprophet is as follows:
>>>>>>>>>
>>>>>>>>> *G(0.000)G(0.000)M(0.000)N(0.000)G(0.000)A(0.000)S(0.000)P(0.000)P(0.000)T(0.000)S(0.000)Q(0.000)K(0.000)*
>>>>>>>>>
>>>>>>>>> Is a probability always 0 somehow an expected behavior? How do I
>>>>>>>>> interpret this? Am I right to say that the 0.9869 Da shift cannot be
>>>>>>>>> localized to any residues?
>>>>>>>>> My concerns comes from the fact that even if the mod cannot be
>>>>>>>>> localized, I would have expected to get a probability of (~0.008, i.e.
>>>>>>>>> 100/length of the peptides) for each residues...
>>>>>>>>>
>>>>>>>>> --
>>>>>>>>> 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 [email protected].
>>>>>>>>> To post to this group, send email to [email protected].
>>>>>>>>> Visit this group at https://groups.google.com/grou
>>>>>>>>> p/spctools-discuss.
>>>>>>>>> For more options, visit https://groups.google.com/d/optout.
>>>>>>>>>
>>>>>>>>
>>>>>>>> --
>>>>>>> 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 [email protected].
>>>>>>> To post to this group, send email to [email protected].
>>>>>>> Visit this group at https://groups.google.com/group/spctools-discuss
>>>>>>> .
>>>>>>> For more options, visit https://groups.google.com/d/optout.
>>>>>>>
>>>>>>
>>>>>> --
>>>>> 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 [email protected].
>>>>> To post to this group, send email to [email protected].
>>>>> Visit this group at https://groups.google.com/group/spctools-discuss.
>>>>> For more options, visit https://groups.google.com/d/optout.
>>>>>
>>>>
>>>> --
>>> 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 [email protected].
>>> To post to this group, send email to [email protected].
>>> Visit this group at https://groups.google.com/group/spctools-discuss.
>>> For more options, visit https://groups.google.com/d/optout.
>>>
>>
>> --
> 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 [email protected].
> To post to this group, send email to [email protected].
> Visit this group at https://groups.google.com/group/spctools-discuss.
> For more options, visit https://groups.google.com/d/optout.
>

-- 
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 [email protected].
To post to this group, send email to [email protected].
Visit this group at https://groups.google.com/group/spctools-discuss.
For more options, visit https://groups.google.com/d/optout.

Reply via email to