TargetLynx / QuanLynx integration issues

Discussions about GC-MS, LC-MS, LC-FTIR, and other "coupled" analytical techniques.

13 posts Page 1 of 1
Dear colleagues
I don't understand the logic behind Quan/TargetLynx integration parameters. I am struggling for years and get more or less acceptable results, but I hate not having full control over the integration especially of small peaks.
Please have a look at the attached chromatogram to see what I mean.

Problem 1: Fictional baseline
Image
https://ibb.co/B4c2rRW

Problem 2: Baseline separation or not?
Image
https://ibb.co/hspGJjw

How can I tell the algorithms to draw the baseline of the peak not from a fictional baseline but from the actual measured data points? How to discriminate what is baseline separated and what not?
The same issues appear using "traditional" and "apex track" integration.
Many thanks in advance,
Jörg
Those peak boundaries - are they manually set or automatically detected? Let's start with automatically detected boundaries. I don't know which options your software exposes so my answer is based on general understanding of ApexTrack. It's easier to see with the 2nd chromatogram that you have a cluster of 2 peaks (they meet at p2):
ImageBecause these peaks are not baseline-resolved ApexTrack determines a shared baseline for the whole cluster between p1 and p3. Hence your Peak1 baseline is below p2 - it's where the shared baseline is located.

If you want the baseline to touch p2 you need to ensure the peaks don't form a cluster:
1. Either Peak1 needs to end earlier (increase Touchdown threshold)
2. Or Peak2 needs to start later (increase Liftoff threshold)
3. Or both of the above.

Alternatively you may try to make the software ignore Peak2:
1. Increase Curvature threshold (Peak2 isn't sharp, so it will be ignored if the threshold is high enough)
2. Or increase the smoothing (increase window size).

But. If the peak boundaries are not autodetected (if you set them manually) then there's yet another possibility - both Peak1 and Peak2 are considered as one peak. Which means too much smoothing is happening and you need to decrease window size. Increasing Touchdown/Liftoff thresholds also may help.
Software Engineer at elsci.io (my contact: stanislav.bashkyrtsev@elsci.io)
sbashkyrtsev wrote:
If you want the baseline to touch p2 you need to ensure the peaks don't form a cluster:
1. Either Peak1 needs to end earlier (increase Touchdown threshold)
2. Or Peak2 needs to start later (increase Liftoff threshold)
3. Or both of the above.

Alternatively you may try to make the software ignore Peak2:
1. Increase Curvature threshold (Peak2 isn't sharp, so it will be ignored if the threshold is high enough)
2. Or increase the smoothing (increase window size).

But. If the peak boundaries are not autodetected (if you set them manually) then there's yet another possibility - both Peak1 and Peak2 are considered as one peak. Which means too much smoothing is happening and you need to decrease window size. Increasing Touchdown/Liftoff thresholds also may help.

The Apex track parameters offered by Waters are VERY limited, see screenshot:
Image

The traditional algorithm offers some more parameters:
Image


The best peak detection control is offered by Chromeleon, by the way...
Your first situation has arisen because of the dip at 0.32min. The software has taken this as base-line and assumed that your peak is actually part of a string of improperly separated things from 0.32min onwards. The truth of the matter is that if you have a very wobbly baseline, then the result is liable to error no matter how sophisticated the integration algorithm.
If the dip at 0.32min is repeatable, you might just be able to deal with it by setting the integration extent in TargetLynx so that it only starts integrating after the 0.32min problem. This is an additional parameter in a TargetLynx method that is not found on the integrator settings part of chromatogram processing (in fact it usually causes trouble, because people get their integration looking good outside TargetLynx, then they process everything, and it's different. The reason is that TargetLynx didn't integrate the entire chromatogram, instead for economy it did only the region around the expected retention time, but if this began half way through a spurious peak, it can get confused).

Your second situation is a matter of personal choice. The integrator believes that there is a second peak (so do I) and if it drew a baseline to touch the data curve, it would be throwing away a genuine, wedge-shaped piece of real peak. The option is there, to make it ignore the drop-line (it's the join-valleys setting). But again, you might end up with a less accurate result than the one that looks strange.

On the whole, there are two things worth remembering about integrators: (1) how good they are is not proportional to how many settings they have; (2) most of us work best with familiar tools; the better we know an integrator, the better they are. I couldn't say what the best integrator is. I've used quite a few, and find that each has its own characteristics. On the whole, when they don't do a good job, it's because there is something a bit wrong with the chromatogram; quite often, this means there is no definitive 'right' behavior. All one can aim for is consistency (assuming the problem can't be solved by changing method/samples etc.)
I have a similar issue: uhplc with msms and sheduled mrm aqusition. The chromatography peak rt is very reproducible but one peak has some co-eluting compounds causing noise around the peak baseline. I wish to integrate everything between two set timepoints down to y-axxis zero. I have not been able to make a targetlynx method for this, it always tries to make some baseline from the chromatogram data, and does this in an inconsistent way. For scheduled mrm the baseline region may be to short for the program to make a good estimate when there is co-eluting minor components causing baseline ripple.

But It should be a simple instruction to set start-stopp time and use zero on y-axxis as baseline on order to sum all components in that window. Can this simple task be done?
The parameter you're looking for, per_oxid, is not in the integrator at all, it's on the integrator page (I think!) of TargetLynx, towards the bottom. There is a bit there where you specify the width of the integration window. It's not start-stop I'm afraid, but a window centred on the peak. In the worst case scenario, simply set it to an enormous value (I think it defaults to only 1 or 2 minutes at most) so that TargetLynx works with the entire chromatogram. Then you will get the same results in TargetLynx as you do when you integrate using the Process menu in MassLynx, and unless your method actually starts collecting data half way through a baseline disturbance, TargetLynx will assess the baseline in a sensible place.
There are two parameters working together, as I found out after struggling quite long (the help function, ans also the QuanLynx tutorials are NOT helpful in explaining any of the non-obvious parameters):
Set the "retention time window" to a relativly narrow value where you expect the peak maximum, it will prevent QuanLynx from integrating the wrong peak. Then set the "window extend" value to a larger number. From my understanding the "retention time window" is multiplied by this number and delimits the range where integration is applied.
@lmh: Indeed, this parameter is centered to the middle of the "retention time window", which is very sub-optimal.
Yes, Target Lynx calls it "Integration Window Extent". The help file's comment is "Extends the retention time window for integration purposes." which is as useful as a chocolate teapot. All these years, I thought it was the size of the region of chromatogram that gets integrated, given in minutes - I didn't know it was a multiplier applied to the retention time window (embarrassing!).

The two parameters are, quite rightly, in different places. The retention time window is on the compound properties tab (because it's to do with identifying the compound). The integration window extent is on the integration tab, because it dictates how the integrator will be used, even though it's not one of the integrators properties in a strict sense.

Waters would do themselves many favours if they wrote better help-files. Their training courses are good, but very pricey, and the materials you get from a training course are often no substitute for a decent help-file. No matter how good the course, you sometimes need a reference for the itty-bitty details of what a particular parameter actually does.
Thanks for explaining the integration parameters. This will sure help me to find better integration methods. Still a bit dissapointing that the simple method to integrate to zero as baseline cant be user set. With scheduled mrm with many compounds in a short method I often set the aquisition window quite narrow to maximise dwell time for each transitions and thus the software can struggle to find baseline. I am happy to hear that I am not the only one that can find the optimisation of integration a bit hard to understand fully. It can be a bit confusing that settings/labels in chromatogram view are different to the ones in targetlynx. Very nice to have this forum for recieving advise beyond the manual. Thanks for taking the time!
bunnahabhain, I recently recorded a video tutorial on ApexTrack (phew, it's harder than it looks after it's done :)), hope it helps at least with some configuration options in the future.

Though many parameters that TargetLynx/QuanLynx expose (including some of those that were discussed in this thread) don't actually pertain to ApexTrack, at least they're not discussed in their docs. They probably put it all into ApexTrack dialogs because it was more user friendly.
Software Engineer at elsci.io (my contact: stanislav.bashkyrtsev@elsci.io)
That apex track tutorial was great, thanks for doing all that work!
Hi everyone - I've taken the time to review each submission and yet cannot resolve my issue.

The problem at hand involves two closely eluting PAH peaks whereby the baseline of the first peak begins okay but is drawn to the valley. The baseline of the second peak begins in the valley and doesnt quite reach the tailing baseline.

I'd like for the baseline to go across the two peaks and a dropline be inserted between them.

What parameters must be altered to achieve this? I've searched the web and attempted many suggestions yet the integrations wont change.

Sorry for not uploading an image. I am doing this on the fly.

thanks in advance

Equipment
CTC Leap autosampler
Agilent GC 6890
Micromass autospec
Targetlynx 4.1
Not sure I got it right. Do you mean something like this:
Image
If so, you need to decrease liftoff threshold so that 2nd peak starts earlier. Or decrease touchdown threshold to make 2nd peak end later. The end goal is to make peaks meet (right now there is a slight gap between them).

If your situation is different, please post a picture. It's hard to read your situation from words. Also would be great if you create a separate topic for your problem.
Software Engineer at elsci.io (my contact: stanislav.bashkyrtsev@elsci.io)
13 posts Page 1 of 1

Who is online

In total there is 1 user online :: 0 registered, 0 hidden and 1 guest (based on users active over the past 5 minutes)
Most users ever online was 1117 on Mon Jan 31, 2022 2:50 pm

Users browsing this forum: No registered users and 1 guest

Latest Blog Posts from Separation Science

Separation Science offers free learning from the experts covering methods, applications, webinars, eSeminars, videos, tutorials for users of liquid chromatography, gas chromatography, mass spectrometry, sample preparation and related analytical techniques.

Subscribe to our eNewsletter with daily, weekly or monthly updates: Food & Beverage, Environmental, (Bio)Pharmaceutical, Bioclinical, Liquid Chromatography, Gas Chromatography and Mass Spectrometry.

Liquid Chromatography

Gas Chromatography

Mass Spectrometry