The routine workflow for layman

[Kelvin Chung]

Hello, 

I have only little experience in Mass spect and proteomic, thus, please forgive my ignorance in that area. 

I am now doing a simple task, identification of a purified protein by routine trypsin digestion, and also find the PTM on that protein, and de novo sequencing of another protein in the gel is also needed. 

The digested sample has been sent to another group and analyzed by Sciex Triple ToF 6600. 

I got the data file and try to analyze it. However, I have a problem in the very beginning step, I import the spectrum file and database file, however, I don;t know what file have to be imported as Identification file. I could not see any description about that, would you mind providing some information for my further reading? 

Thank you very much 

Kelvin

[Harald Barsnes]

Hi Kelvin,

Have you had a look at our tutorials: https://www.compomics.be/bioinformatics-for-proteomics?

They should give you the background needed to understand and use our tools.

Note however that our software is mainly created for discovery-based proteomics studies, and may therefore not align all that well with your particular study focusing on a single known protein.

Best regards,

Harald

[Juan Camilo Rojas Echeverri]

Hi,

Just wanted to add to this:

 

" and may therefore not align all that well with your particular study focusing on a single known protein"

I have been using PeptideShaker so far mostly with studies of a model protein considering many PTMs. The protein score is always doubtful since (as expected) it can't estimate accurately a protein FDR distribution.

For the peptides and PTM annottations it's mixed. Peptides are still usually classified (max) as doubtful but with 100 confidence.

For PTMs it can have full classification of "Confident", but in some scenarios even if the MSMS scan is well annottated, it can give low confidence values.

My guess in this particular scenario, is that many of the other MSMS events were so superior with respect to absolute intensities, even this good MSMS that I would happily accept through manual annotation is still considered "bad".

All in all, Kelvin, I think it can work for you as long as you make sure to rely more on manual checks on the MSMS annottations than the statistic filtering when you are working with a single protein. In my humble experience, the annottation and visualization of PeptideShaker is vastly superior from many other tools I have tested, SPECIALLY when you know specific reporter ions or losses from the PTM that you are looking for since it can annottate them (although they are not used for scoring? Am I correct to assume this, Harald?).

Cheers and good luck!

Sincerely,

JC

[Harald Barsnes]

Hi JC,

Thanks for the input! You are of course correct in that PeptideShaker can also be used to look at this type of data if relying more on manual inspection of the spectra and less on the statistical validation.

As for the PTM-related ions being part of the scoring or not, this is mainly up to the individual search engines. Some include this information in their scoring and some do not.

We do consider this information when trying to target the location of the individual PTMs in the peptide sequence though. 

Best regards,

Harald

[Kelvin Chung]

 Harald and JC 

Thank you very much for your reply. I understand the peptide shaker is a much more powerful tools not design for simple single protein analysis, but the interface is so friendly and good looking, therefore I try to using it for my routine work.