ProteinProphet output number of unique peptides vs total number of peptides

[muriell...@gmail.com]

Hi all,

I have two questions concerning the output of ProteinProphet :

1) Why do I have some entries for which the total number of peptides is lower than the number of unique peptides?  (see columns 5 and 6 below, an extract of my output table)

2) Does the column "total number of peptides" correspond to spectral counts?

Thank you very much ,

Murielle

[David Shteynberg]

Hi Murielle,

Thanks for using the TPP and submitting your question here.   When I try the latest version of the TPP and run ProteinProphet through there, the columns exported in my tab separated file from ProteinProphet are different from yours.  Can you please provide a bit more information about how you generated this file and which version of the software you used? 

Cheers,

-David

--
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-discu...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/spctools-discuss/be3b35c9-e308-4e35-8551-03425f7d1704n%40googlegroups.com.

[muriell...@gmail.com]

Hi David,

Thank you for such a quick reply!

I used tpp version 6.2.0 run via a singularity image and I produced the table by including the EXCELXX option in proteinprophet.

Cheers,

Murielle

[David Shteynberg]

Hi Murielle,

Sorry this is a little confusing.  Column "num unique peps" represents total number of unique peptides mapping to the protein.  The "tot num peps" counts the PSMs that are "contributing evidence" (last column) to the protein, sometimes this number can be lower because not all peptides mapping are contributing evidence and spectral counts can be low per peptide.  

The answer to your second question is yes, the "tot num peps" is the spectral count for "is contributing evidence" peptides.

Cheers,

-David

To view this discussion on the web visit https://groups.google.com/d/msgid/spctools-discuss/04691060-aedc-4d65-87a0-ae40aa92f889n%40googlegroups.com.

[Hampton, Brian]

Hello David,

I have difficulty understanding this.  I had always operated under the assumption that the "num unique peps" would be less than or equal to the "tot num peps" - because - I thought "tot num peps" = "num unique peps" + number of (shared) peptides also mapping to given protein.  Is this not the case and if so how would "tot num peps" ever be less than "num unique peps"?

Best regards,

Brian

Brian Hampton
Proteomics Core Lab

Center for Vascular and Inflammatory Diseases

University of Maryland School of Medicine

BioPark One Rm 307

800 West Baltimore Street

Baltimore, MD. 21201

(410)706-8207

---

From: 'David Shteynberg' via spctools-discuss <spctools...@googlegroups.com>
Sent: Wednesday, June 28, 2023 11:33 AM
To: spctools...@googlegroups.com <spctools...@googlegroups.com>
Subject: Re: [spctools-discuss] ProteinProphet output number of unique peptides vs total number of peptides

 

To view this discussion on the web visit https://groups.google.com/d/msgid/spctools-discuss/CAGJJY%3D9cFy1deU1TrZzMRyJ7GqTX4ZX-bByL6GDFj6ZEcU6Kug%40mail.gmail.com.

[Emma Whittington]

I am also a little confused/requiring clarification. For label-free quantification would you use the "tot num peps" for spectral counts when calculating something like NSAF? From your description, it sounds like that's the total number of contributing spectra.  

all the best, 

Emma 

[David Shteynberg]

Hello Brian,

Thanks for the discussion.  The way this is possible is because there are degenerate peptides that map to multiple proteins, in which case only those that have sufficient weight will be counted as "is contributing evidence" peptides.  Only those PSM  instances that are contributing evidence are counted in the "tot num peps" value.   Please let me know if further clarification would be helpful or if you have other questions.

Cheers,

-David

To view this discussion on the web visit https://groups.google.com/d/msgid/spctools-discuss/BL0PR03MB4259C80B5A47840AFEEB12948924A%40BL0PR03MB4259.namprd03.prod.outlook.com.

[David Shteynberg]

Hi Emma,

Yes, for spectral counting of proteins it seems appropriate to consider only "is contributing evidence" peptides,  "tot num peps"  would be the value to use.

Cheers,

-David

To view this discussion on the web visit https://groups.google.com/d/msgid/spctools-discuss/f5e7ab0f-7f2a-4494-8560-652629ef9812n%40googlegroups.com.

[Hampton, Brian]

Hi David, thanks for the quick reply.

I think I see now.  When there is some ambiguity which prevents determining a precise peptide sequence, for example incomplete fragment ion series, or isobaric AA assignments etc. then there must be a probability calculation performed, the result of which attempts to assign this peptide to the more likely protein candidate.  The result of which is recorded in the "total num peptides" value.  In the case of a tie, is the value added to all possible protein hits or not used?

Thanks a bunch!

Brian

---

From: 'David Shteynberg' via spctools-discuss <spctools...@googlegroups.com>

Sent: Wednesday, June 28, 2023 12:37 PM

To view this discussion on the web visit https://groups.google.com/d/msgid/spctools-discuss/CAGJJY%3D8eH7q7L7moDYWT7vgmgJrFtJpdyyS4ZZbuX4Ygy9utuQ%40mail.gmail.com.

[David Shteynberg]

Hi Brian,

I just confirmed in the code that it is counted here when the weight of the peptide is minimum 0.5 and the probability minimum 0.2.

Cheers,

-David

To view this discussion on the web visit https://groups.google.com/d/msgid/spctools-discuss/BL0PR03MB42598ECFAC9E9948FA1D53908924A%40BL0PR03MB4259.namprd03.prod.outlook.com.