False identified transitions and product ions list

support
False identified transitions and product ions list William  2014-10-08 21:13
 
Hi Brenda,
I imported data file of CE optimization (Agilent)and result as shown in the attachment. There are two questions:
1. As a practice, I added 4 synthetic peptides only but turned out identifying almost all 14 peptides list (green dot). But when I expanded those false identified transitions, it is actually noise level. So how can I set up a thershold noise level to avoid so many false idetifications?
2. As you can see that I put product ions of "m/z>precursor" to "3 ions" in the filter, but it seems to still pick up product ions with m/z < precursor. Is there a fix for it?

Thanks,
William
 
 
Brendan MacLean responded:  2014-10-08 21:28
Hi William,
This is a pretty small list, and fairly obvious when there is not a peak. I would just use manual inspection to determine when you are measuring a peptide and when not. Otherwise, see the Advanced Peak Picking Models tutorial, if you actually want to create a scoring model that would give you a q value for this, and probably a poor one for the transitions you are showing in your screenshot. Seems like introducing unnecessary complexity at this point to me, though.

You are most likely seeing your filter values not applied, because you have not indicated they should be applied in the Transition Settings - Library tab. If you use "From filtered ion charges and types" then your precursor m/z filter will not be applied. You need to use "From filtered product ions" instead, though this also makes no sense with a filter that includes only "3 ions". If you really want to use such a restrictive filter, then you should just un-check the checkbox "If a library spectrum is available, pick its most intense ions", which is clearly checked, and finding library spectra, based on your screenshot.

Hope this helps. Good luck with your CE optimization.

--Brendan
 
William responded:  2014-10-08 21:58
Thanks for response. I used the small list as a practice. However, our usual list is in a range of 40-50 peptides in which manually check each one will be time intensive. Just find the new tutorial of "Advanced Peak Picking Model", but wonder it will be more convenient (user friendly)to have a threshold option for easy control esp during CE optimization stage. Maybe worth consider in next release?

Thanks,
William
 
Brendan MacLean responded:  2014-10-08 22:00
Sure. It is on our list:

https://skyline.gs.washington.edu/labkey/issues/home/issues/details.view?issueId=348

Thanks for your feedback.

--Brendan