Menu
▾
▴
Re: [Psidev-pi-dev] mzid 1.2 reviews
|
From: Binz Pierre-A. <Pie...@ch...> - 2017-01-19 08:51:37
|
Hi all, Yep, we are redundantly coming to thosse definitions and interpretations of what those peptidic things are... A peptide is, according to IUPAC gold book https://goldbook.iupac.org/P04479.html: Peptides : Amides<https://goldbook.iupac.org/A00266.html> derived from two or more amino carboxylic acid molecules (the same or different) by formation of a covalent bond<https://goldbook.iupac.org/C01384.html> from the carbonyl carbon of one to the nitrogen atom of another with formal loss of water. The term is usually applied to structures formed from α-amino acids, but it includes those derived from any amino carboxylic acid. Therefore it includes a combination of all 20 amino acids + all kinds of modifications A peptide is not a peptide ion in the sense of a MS ion as it is usually the charge neutral representation of a combination of a-aminoacids in the case of protein peptides What we represent here are either the molecular ion (peptide sequence represented by one-letter codes + optional PTMs + charged aduct (H+) + charge state) which we can call peptide ion : [ANSpK +2H] 2+ (where Sp is a phosphoserine) which we can interpret with only the sequence with PTMs ANSpK or only with the primary sequence with no PTM ANSK My suggestion then is to keep the term peptide as the most generic and does not need to specify whether it carries PTMs or not: name: peptide-level spectral count # of ANSpK name: sequence-level spectral count # of ANSK name: peptide ion-level spectral count # of [ANSpK+2H]2+ Now if you feel that the term “peptide-level” will be ambiguous (people will not know that peptide sequence and peptide (with or without mods) are different, we can give a clear definition with example or I could recommend using “peptide form level” or “peptide species level” as analogy to protein form or protein species vs protein (top-down inspired definition), but this adds a new definition and make our filkes larger for no good reasons. Pierre-Alain De : Juan Antonio Vizcaino [mailto:ju...@eb...] Envoyé : mercredi 18 janvier 2017 18:23 À : Eric Deutsch Cc : Tobias Ternent; psi...@li...; Robert Chalkley; le...@im...; jul...@ru...; Ghali, Fawaz; mar...@ru...; mayerg97; Andy Jones; Perkins, Simon; psi...@li...; ju...@ed...; Eugen Netz; Harald Barsnes Objet : Re: [Psidev-pi-dev] mzid 1.2 reviews Fine for me. Sorry, I did not realise about what Eric mentioned. He is completely right. Cheers, Juan On 18 Jan 2017, at 17:18, Eric Deutsch <ede...@sy...<mailto:ede...@sy...>> wrote: Hi everyone, I agree that splitting this into 3 terms is good. I disagree that we should misuse the term “peptide” and then try to clarify it with a parenthetical expression. I don’t think this is good controlled vocabulary practice. I suspect the true ontology devotees would be horrified by this practice. May I suggest that we use the style that I adopted for the mzIdentML CV reorg? This keeps the term name concise, and we should rely on the definition to remove any ambiguity. I suggest these names instead: name: distinct peptide sequence-level spectral count name: modified peptide sequence-level spectral count name: peptide ion-level spectral count What do you think? Does this capture the desired concepts clearly? Regards, Eric From: Juan Antonio Vizcaino [mailto:ju...@eb...<mailto:ju...@eb...>] Sent: Wednesday, January 18, 2017 1:42 AM To: Eric Deutsch <ede...@sy...<mailto:ede...@sy...>> Cc: mayerg97 <ger...@ru...<mailto:ger...@ru...>>; Andy Jones <And...@li...<mailto:And...@li...>>; psi...@li...<mailto:psi...@li...>; Perkins, Simon <Sim...@li...<mailto:Sim...@li...>>; Harald Barsnes <Har...@ui...<mailto:Har...@ui...>>; Yasset Perez-Riverol <yp...@eb...<mailto:yp...@eb...>>; Tobias Ternent <to...@eb...<mailto:to...@eb...>>; jul...@ru...<mailto:jul...@ru...>; mar...@ru...<mailto:mar...@ru...>; lfi...@st...<mailto:lfi...@st...>; ju...@ed...<mailto:ju...@ed...>; Eugen Netz <eug...@tu...<mailto:eug...@tu...>>; Mathias Walzer <wa...@in...<mailto:wa...@in...>>; Oliver Kohlbacher <oli...@un...<mailto:oli...@un...>>; le...@im...<mailto:le...@im...>; Robert Chalkley <cha...@cg...<mailto:cha...@cg...>>; Ghali, Fawaz <F....@li...<mailto:F....@li...>>; Salvador Martínez de Bartolomé <sma...@pr...<mailto:sma...@pr...>>; psi...@li...<mailto:psi...@li...> Subject: Re: mzid 1.2 reviews Hi all, I agree with Eric on this. The more clarity the better. What about? [Term] id: MS:1001842 name: peptide PSM count (peptide as the raw peptide sequence) def: "The number of MS2 spectra identified for a given peptide (considering the raw peptide sequence, and merging all charge states, modifications in the same peptide sequence are not considered separately) in spectral counting." [PSI:PI] xref: value-type:xsd\:int "The allowed value-type for this CV term." is_a: MS:1001805 ! quantification datatype is_a: MS:1002015 ! spectral count peptide level quantitation [Term] id: MS:1001xxx name: peptide PSM count (peptide as the combination between peptide sequence and modifications) def: "The number of MS2 spectra identified for a given peptide (considering the peptide sequence plus modifications, and merging all charge states) in spectral counting." [PSI:PI] xref: value-type:xsd\:int "The allowed value-type for this CV term." is_a: MS:1001805 ! quantification datatype is_a: MS:1002015 ! spectral count peptide level quantitation [Term] id: MS:1001xxx name: peptide PSM count (peptide as the combination between peptide sequence, modifications and charge state) def: "The number of MS2 spectra identified for a given peptide (considering the peptide sequence and modifications, and one single charge state) in spectral counting." [PSI:PI] xref: value-type:xsd\:int "The allowed value-type for this CV term." is_a: MS:1001805 ! quantification datatype is_a: MS:1002015 ! spectral count peptide level quantitation Of course more variants would need to be created if we would want to mix the definitions above with different charge states, but I think these three definitions would be ok. What happens also in the case of ambiguity in the modification position? So, it could be even more complex. But I think this level is probably ok for the current state of things. Cheers, Juan On 17 Jan 2017, at 15:08, Eric Deutsch <ede...@sy...<mailto:ede...@sy...>> wrote: This is pretty good, but I think we should make it completely clear whether it is the number of PSMs per peptide sequence, per peptide with modifications, or per peptide ion (including separate charge states). The plain word peptide is often used as a standin for any of these 3 concepts.. Thanks, Eric From: mayerg97 [mailto:ger...@ru...<mailto:ger...@ru...>] Sent: Tuesday, January 17, 2017 4:39 AM To: Jones, Andy; psi...@li...<mailto:psi...@li...> Cc: Juan Antonio Vizcaino (ju...@eb...<mailto:ju...@eb...>); Perkins, Simon; 'Harald Barsnes'; 'Yasset Perez-Riverol'; to...@eb...<mailto:to...@eb...>; jul...@ru...<mailto:jul...@ru...>; mar...@ru...<mailto:mar...@ru...>; lfi...@st...<mailto:lfi...@st...>; ju...@ed...<mailto:ju...@ed...>; Eugen Netz; Mathias Walzer; Oliver Kohlbacher; le...@im...<mailto:le...@im...>; 'Robert Chalkley'; Ghali, Fawaz; sma...@pr...<mailto:sma...@pr...>; 'Eric Deutsch; psi...@li...<mailto:psi...@li...> Subject: Re: mzid 1.2 reviews Hi Andy, yes we did a CV restructuring, see https://github.com/HUPO-PSI/mzIdentML/issues/25, but I think this problem is not related to the restructuring. What about [Term] id: MS:1001842 name: peptide PSM count def: "The number of MS2 spectra identified for this peptide in spectral counting." [PSI:PI] xref: value-type:xsd\:int "The allowed value-type for this CV term." is_a: MS:1001805 ! quantification datatype is_a: MS:1002015 ! spectral count peptide level quantitation Best regards, Gerhard Am 11.01.2017 um 16:56 schrieb Jones, Andy: Hi all, (Sent to PSI-PI mailing list plus authors of the mzid paper, in case any of you are not on the mailing list). We have received 5 reviews of the mzid 1.2 specifications via the PSI process, as well as the 2 reviews received on the manuscript from MCP, for which it makes sense to consider together. I think most of these can be addressed fairly simply. I have already fixed a few typos in the specification document that were identified, and made one or two other sensible clarifications to wording. These changes have been checked into GitHub. I have started response documents to both processes in the relevant zip files. If you would like to contribute, please could you send me any feedback within one week (by 19th Jan) if possible. After that point, I will submit the mzid 1.2 specs back to the PSI process (for which Sylvie the editor recommended minor corrections), unless anything comes up that needs extensive further discussion. Once they are accepted, I will resubmit the paper to MCP, as reporting the mzid 1.2 final specifications. Just flagging a few points for specific people to address: - Robert and Alexander: one of the reviewer asks whether your cross-linking software will export mzid 1.2, anything you can say here? - Fawaz – one of the reviewers seemed to find some validation errors with some of our mzid 1.2 files, please could you take a look - Juan – one of the MCP reviewers suggests we change the figures. Can you give me a view on whether you think we need to do this? I’m of the view to make minor improvements, but not radically change them. - Gerhard – see response to MCP reviews. One of the reviewers would like to know how spectral counts can be added at the peptide-level. I think this may need a new parent term for one given CV term (see my notes in the response to reviews doc), could you take a look? Best wishes Andy -- -------------------------------------------------------------------- Dipl. Inform. med., Dipl. Wirtsch. Inf. GERHARD MAYER PhD student Medizinisches Proteom-Center DEPARTMENT Medical Bioinformatics Building ZKF E.049a | Universitätsstraße 150 | D-44801 Bochum Fon +49 (0)234 32-21006 | Fax +49 (0)234 32-14554 E-mail ger...@ru...<mailto:ger...@ru...> www.medizinisches-proteom-center.de<http://www.medizinisches-proteom-center.de/> |
