Issue 2: Support second round CE optimization

Status:open
Assigned To:Kaipo Tamura
Type:Todo
Area:Skyline
Priority:4
Milestone:4.3
Opened:2011-07-07 08:15 by Brendan MacLean
Changed:2017-09-21 14:14 by Kaipo Tamura
Resolved:
Resolution:
Closed:
2011-07-07 08:15 Brendan MacLean
Title»Support second round CE optimization
Assigned To»Brendan MacLean
Type»Defect
Area»Skyline
Priority»3

2011-07-07 08:15 Brendan MacLean
Several people have noted that it is difficult to re-optimize transitions for which the optimization based on the linear equation failed to capture an apex in the CE dimension. It should be made possible to run subsequent rounds of optimization including only the precursors for which an optimal value was not clearly measured.

This is a little complicated due to the need to know the central CE value from which the optimization range was generated. Currently that central value is assumed to be the one predicted by the current linear equation for CE in the document settings. In order to make this second round optimization work, Skyline might need to snapshot best measured CE values for certain precursors into the document settings, to be correlated with the CE optimization data during import.

We could certainly make this work, and it would be a great improvement to the CE optimization functionality. Unfortunately, it is not a top priority at the moment.


Steve S., who originally requested this also adds the ideas below for an future round of development on CE optimization.

One final note: if you do at some point implement more CE functionality, it would be nice to also
-read in the CE from raw files, to allow previously optimized CE values (from other software/manual optimization) to be stored in Skyline.
-have the option to display individual CE values when viewing data in skyline (alongside the fragment mass, intensity, retention time).

2011-07-07 08:29 Brendan MacLean
TypeDefect»Todo

2014-04-12 15:02 Brendan MacLean
Assigned ToBrendan MacLean»bspratt
Not sure when we might have time for this, but a great way to do it would be to allow the inclusion of values in a CE optimization library (which you are going to implement) so that the seed CE values that get used come from the library when available, and from a linear equation on precursor m/z when no library entry is available.

2014-08-28 14:53 Brendan MacLean
Assigned Tobspratt»kaipot@u.washington.edu
Milestone»3.1
This is pretty tricky to get working in the same way that we do the first round, or by using values in a library without relying on the actual CE values that the instrument measured. It is one thing to connect CE values with a predictive equation that will stay the same between when the user exports their method and then imports the data, and completely different to make this more complex and relying on values that could change, like values in a library file.

We will need to work something out, though. AB SCIEX is requesting that we support multiple-iteration optimization of ion mobility settings for their Solexa filter, and that should be very similar.

2015-11-07 12:28 Brendan MacLean
Priority3»4
Milestone3.1»

2017-09-21 14:14 Kaipo Tamura
Assigned ToGuest»kaipot@u.washington.edu
This is now sort-of becoming possible, since Brendan just fixed CE optimization for small molecules to work with Explicit Collision Energy. In theory, someone could do a second round by entering the values from the first round into Explicit Collision Energy. Or we could allow some way of using what is in the optimization library... Or something like we are now doing for compensation voltage.

This has not been a very heavily requested feature, though.