signal to noise calculation

support
signal to noise calculation xue shi  2022-04-05
 

Hi Skyline team,
I am trying to calculate signal to noise. Does skyline have signal to noise calculation option?
I also found skyline has background area calculation.
https://skyline.ms/wiki/home/software/Skyline/page.view?name=tip_peak_calc
However, I don't know how skyline calculate background area for this case. Please find attachment for the figure.
Could you please let me know which choice skyline to calculate the s/n?
Thank you,
Xue

 
 
Nick Shulman responded:  2022-04-05
In the document that you attached, the lower picture shows the area that Skyline would consider the background.
After the integration boundaries have been chosen, Skyline looks at the intensity of the chromatogram at the peak start time and peak end time. The lower of those two intensities is the background level, and the background level times the width of the peak is (usually) equal to the background area.

(If the chromatogram happens to dip below the background level in between the peak boundaries, then the background area would end up being less than background level times width).

There are many types of noise which affect your mass spectrometry measurements. The background area that Skyline gives probably represents other stuff that is in the mass spectrometer which contributes to the observed signal.

Other types of noise are "shot noise" which is noise that you always get when you are sampling a finite number of molecules. Shot noise has something to do with Poisson Distributions is always proportional to the square root of the number of ions of interest that were inside of the mass spectrometer when the measurement was made. You can figure out how many ions were actually measured by looking at the automatic gain control (AGC) values of the spectra. Skyline does not have any features that make it easy to figure out what the AGC was, but your own mass spectrometry software might help there.

-- Nick
 
xue shi responded:  2022-04-05
Hi Nick,
Thank you for prompt response. so will you think the area divided by background area as signal to noise ratio or not?
Thanks,
Xue
 
Mike MacCoss responded:  2022-04-05
In most cases the reciprocal of the coefficient of variation is the best way to represent S/N. Where signal is the mean and noise is the SD. This is the only way to account for shot noise. Other methods that just use the standard deviation of the background or blank only consider the Johnson noise which is a minor portion of the overall measurement noise.

Anyway, just something to consider. If you really want a signal to "noise" versus a signal to "background"

-Mike