smith-chem-wisc / FlashLFQ

Ultra-fast label-free quantification algorithm for mass-spectrometry proteomics
GNU Lesser General Public License v3.0
19 stars 14 forks source link

Caution messafe in running #57

Closed Dmorgen closed 5 years ago

Dmorgen commented 6 years ago

Hi,

I've recieved the following message - "Caution! PSM with sequence **** could not be read; a peptide with the same modified sequence but with different monoisotopic mass has already been added

I'm guessing itis the consolidation of PSMs to a single peptide?

Thanks! David.

rmillikin commented 6 years ago

Hi David,

The monoisotopic mass in the identification file should be the theoretical mass including the mass of any modifications, so there should only be one mass per modified sequence. This is required to calculate accurate ppm errors, mainly. If a sequence has more than one mass it can confuse some calculations. So yep it does have to do with consolidating PSMs into peptides.

Dmorgen commented 6 years ago

Thanks! I'm using Byonic, which doesn't provide a calculation of the theoretical mass. do you know of any tool that will do that for a peptide list?

Thanks! David.

rmillikin commented 6 years ago

Hmm. Interpreting amino acid sequences is straightforward and available in many places, e.g. protein prospector. The hard part is interpreting the modifications, since every program has a different way of annotating those. That's why I require that the user put in the monoisotopic mass, because I want to support different search inputs but it would take an enormous amount of effort to figure out how everyone annotates their modifications.

I guess you would need a Byonic -> mass converter specifically, and I don't know if any exist, unfortunately.

I suppose as a crude stop-gap you could take the experimental masses for each unique sequence and just use the average of those experimental masses as the theoretical. That would get you up and running but would introduce another source of error...

trishorts commented 6 years ago

if you took .mzid input, you could get mods in a standard format. I guess there are some data converters to .mzid

rmillikin commented 6 years ago

I think even then it would be very difficult because of the nomenclature differences. mzID would help for where the mod is localized to, and how to read the name, etc, but different programs have different names for the same mod. MetaMorpheus and MaxQuant have different names for phosphorylation, for example. This makes interpreting the chemical formula of the mod very difficult, unless the chemical formula is annotated in the mzID.

Dmorgen commented 6 years ago

I've back calculated the theoretical masses from the observed mass and ppm. does the mass have to be M+H or M?

Thanks! David.

rmillikin commented 6 years ago

The mass needs to be the neutral peptide mass (M, not M+H)

Dmorgen commented 6 years ago

Thanks!

From: rmillikin [mailto:notifications@github.com] Sent: Tuesday, May 29, 2018 5:26 PM To: smith-chem-wisc/FlashLFQ FlashLFQ@noreply.github.com Cc: David Morgenstern david.morgenstern@weizmann.ac.il; Author author@noreply.github.com Subject: Re: [smith-chem-wisc/FlashLFQ] Caution messafe in running (#57)

The mass needs to be the neutral peptide mass (M, not M+H)

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHubhttps://github.com/smith-chem-wisc/FlashLFQ/issues/57#issuecomment-392796243, or mute the threadhttps://github.com/notifications/unsubscribe-auth/AiHaodpYTiWMXNsr0LQyAnJ4beWEyuBPks5t3VqBgaJpZM4UPT7r.