NSAF reported by ProteinShaker

[William Stone]

I am just starting to look at the output from ProteinShaker and notice "M2 quant." values like 2.74E03 (= 2740).  I  thought NSAF values run from zero to one?   What am I missing? 

[Marc Vaudel]

Hi William,

NSAF values are the number of (validated) spectra divided by the protein length, so there is no reason for them to be bounded to 1 :-) Note that in PeptideShaker, the number of spectra for a peptide is weighted by the number of times a peptide is occurring in a validated protein (if a peptide is found in 3 validated protein groups, the each spectrum counts 1/3), and the protein length only includes sections of the protein likely to generate measurable peptides based on the distance between two cleavage sites.

In addition, by default the NSAF values are scaled so that the sum of all proteins (or rather the leading protein of all validated protein groups) makes a mass of 2 ug, and the MS2 Quant value is then given in fmol. Note that you can change the scaling using the Edit -> Project Preferences menu, and set there the protein amount that you injected and unit of display. You can also change the normalization method or disable it altogether.

Hope it helps, apologies that this is not well documented, we will improve this in the next version,

Marc

Le jeu. 5 déc. 2019 à 15:52, William Stone <drbil...@gmail.com> a écrit :

I am just starting to look at the output from ProteinShaker and notice "M2 quant." values like 2.74E03 (= 2740).  I  thought NSAF values run from zero to one?   What am I missing? 

--
You received this message because you are subscribed to the Google Groups "PeptideShaker" group.
To unsubscribe from this group and stop receiving emails from it, send an email to peptide-shake...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/peptide-shaker/ac6ad8c0-a00f-4f96-8df4-56c97afaf871%40googlegroups.com.

[William Stone]

Thank you, Dr. Vaudel, for your detailed response. I very much appreciate your efforts in the development of the Compomics software suit.  Florens et.al (Methods,2006) state "To account for the fact that larger proteins tend to contribute more peptide/spectra, spectral counts were divided by protein length (L), defining a Spectral Abundance Factor (SAF) [15].  SAF values were then normalized against the sum of all SAFs for a particular run (removing redundant proteins) allowing us to compare protein levels across different runs ....Normalized Spectral Abundance Factors (NSAFs) were calculated as follows [21]: (2).  NSAFk= (SpC/L)k/sum i+1 to N (SpC/L)i.  Where SpC= spectral counts and L=protein length as measured by the number of amino acid residues. NSAF values should range from 0 to 1, with values closer to 1 indicating higher protein levels. "

I will try to see if the protein shaker project options allow this type of output. 

Thanks, Bill 

On Thursday, December 5, 2019 at 4:31:33 PM UTC-5, Marc wrote:

Hi William,

NSAF values are the number of (validated) spectra divided by the protein length, so there is no reason for them to be bounded to 1 :-) Note that in PeptideShaker, the number of spectra for a peptide is weighted by the number of times a peptide is occurring in a validated protein (if a peptide is found in 3 validated protein groups, the each spectrum counts 1/3), and the protein length only includes sections of the protein likely to generate measurable peptides based on the distance between two cleavage sites.

In addition, by default the NSAF values are scaled so that the sum of all proteins (or rather the leading protein of all validated protein groups) makes a mass of 2 ug, and the MS2 Quant value is then given in fmol. Note that you can change the scaling using the Edit -> Project Preferences menu, and set there the protein amount that you injected and unit of display. You can also change the normalization method or disable it altogether.

Hope it helps, apologies that this is not well documented, we will improve this in the next version,

Marc

Le jeu. 5 déc. 2019 à 15:52, William Stone <drbil...@gmail.com> a écrit :

I am just starting to look at the output from ProteinShaker and notice "M2 quant." values like 2.74E03 (= 2740).  I  thought NSAF values run from zero to one?   What am I missing? 

--
You received this message because you are subscribed to the Google Groups "PeptideShaker" group.

To unsubscribe from this group and stop receiving emails from it, send an email to peptide...@googlegroups.com.

[Marc Vaudel]

Hi again,

Thanks for your answer and the kind words on our work! If you want to normalize to the sum as done in the original publication please select "Relative" in the normalization setting, the NSAF index should then be normalized to the sum and available in percent or ppm. If something does not work please let us know.

Best regards,

Marc

To unsubscribe from this group and stop receiving emails from it, send an email to peptide-shake...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/peptide-shaker/2a7c79e8-4241-43a0-8889-003fe162a748%40googlegroups.com.