Which formula to choose for combining peptides intensity into protein intensity ?

Which formula to choose for combining peptides intensity into protein intensity ? nicolas pierre  2022-05-05

Hello Skyline Team,

I have measured endogenous peptides with their heavy counterpart by SRM. I have also spike a protein in my sample for normalisation.
I have some questions for translated results at protein level.
When I divide the intensity of endogenous peptide by their heavy counterpart it become a ratio and in many cases this ratio totally change the ponderation of each peptide for their contribution to the protein intensity. Typically, peptides representing the highest intensity (the one which would contribute the more to protein signal with the addition of peptide intensities) for the protein can, when divided by its heavy counterpart, be the one with the lowest ratio among the other peptide of the protein. In this case this will totally change the ponderation. I am not very confident with this phenomena because usually signal with the lowest intensity are those with the lowest quality (near LOD, lowest level of precision, lowest quality of the peak...). It means that with a division by the heavy peptides I will give more importance (not systematically) to signal with the lowest quality.
In fact their is different solution to combine the signal of peptide into protein signal. For instance:

-Signal of protein A= (Signal of endogenous peptides 1/ signal of heavy peptide 1) + (Signal of endogenous peptides 2/ signal of heavy peptide 2)
-Signal of protein A = (Signal of endogenous peptides 1 + Signal of endogenous peptides 2)/(signal of heavy peptide 1 + signal of heavy peptide 2)

As explained, the first formula can change the ponderation of the peptides contributing to the protein signal, the second one will not. We could imagine other solutions...Also, I have spiked a protein for normalising my signal and I do not know where I can put it in the formula, where would be the best place..
Do you have some advises for this problem ? In there some Skyline solutions ?
Thank you in advance for your help.

Nicolas Pierre

Nick Shulman responded:  2022-05-05
When Skyline calculates the normalized area of a peptide or a protein, Skyline always uses your second formula there, and divides the sum of the endogenous areas by the sum of the heavy areas. This has the effect of giving greater weight to the more intense peptides.

If you do what you have in the first formula, you actually end up giving the same amount of weighting to the peptide with low signal as the peptide with high signal. There might be valid reasons to want to do that, but Skyline never does that.
Note that, for that first formula, you need to divide everything by two, because you are taking the average of those two ratios, not adding the ratios together.

Skyline will calculate the normalized peptide and protein areas using your second formula, so you do not need to do it yourself unless you are trying to verify that Skyline is getting the right answer.
In the Document Grid, there is a column called "Normalized Area" which will tell you the peptide's normalized area. There is also a column called "Protein Abundance" which will tell you the normalized area of the protein. (There is a binoculars button at the top of the Report Editor which can help you search for any column by name).
The normalization method that Skyline uses is whatever you have chosen at:
Settings > Peptide Settings > Quantification

If you would like to learn more about the Document Grid, this tutorial is helpful:

If you would like to learn more about using Skyline to compare abundances between groups of replicates, you should look at the Grouped Comparison tutorial:

-- Nick
nicolas pierre responded:  2022-05-06
Fantastic response ! Thank you very much.