Gaussian peak fitting

This section allows you to vote for new features that may be implemented into Clarity.
Forum rules
You have to be logged in to be able to vote.
This section lists feature requests with polls. You can:
  • Vote for the feature to increase its priority and discuss how it should be implemented.
  • How to vote? Just log in to your forum account and the voting options will appear.
  • If you have not found your desired feature here, create a new thread for it.
  • You will not be able to create the poll yourself, but after the checking of the suggestion, we will create the poll for you.
We do take this poll seriously, however it is not the only source of priorities for the development.
Post Reply

Do you need a deconvolution mechanism performed by the means of gaussian peak fitting?

Yes.
5
100%
No.
0
No votes
 
Total votes: 5

mkawatasu
Posts: 4
Joined: Wed Jun 03, 2009 11:07 am

Gaussian peak fitting

Post by mkawatasu »

It would be nice if Clarity could have Gaussian peak fitting. This is our customer's request.

Internal issue tracking ID: 36122
Last edited by Daniel Mentlik on Mon Feb 14, 2011 4:31 pm, edited 1 time in total.
Reason: Added a poll
Miyako Kawakatsu

User avatar
Daniel Mentlik
DataApex Support
DataApex Support
Posts: 353
Joined: Fri Mar 27, 2009 3:15 pm

Re: Gaussian peak fitting

Post by Daniel Mentlik »

Dear Miyako,
we have already tried one deconvolution mechanism for these cases in the previous versions of Clarity, but rejected it because it gave insufficient results when used on more complex peak clusters over the course of the whole chromatogram. May I ask you how complex are your expected peak clusters?
Thank you for the answer,
Best Regards,
Daniel
Daniel Mentlí­k
DataApex

mkawatasu
Posts: 4
Joined: Wed Jun 03, 2009 11:07 am

Re: Gaussian peak fitting

Post by mkawatasu »

Hello,
I had the typical chromatogram sent by the user (see the attached file). The sample is low Mw synthesized compound. Two peaks are separated incompletely. The user would like to apply Gaussian fitting to it, and make the simple two peaks. He is using Origin software with Clarity. http://www.originlab.com/index.aspx?go= ... akAnalysis
TEST.JPG
Miyako Kawakatsu

User avatar
Daniel Mentlik
DataApex Support
DataApex Support
Posts: 353
Joined: Fri Mar 27, 2009 3:15 pm

Re: Gaussian peak fitting

Post by Daniel Mentlik »

Dear Miyako,
the chromatogram you have showed us seems more like there are at least four unresolved peaks - the first "major" one has strange lumps both before and after its maximum. I'm afraid that the deconvolution procedure could create any result you wish from such peak cluster, based on the conditions you set to the procedure - it would just be a guessing from the crystal ball. Moreover, peaks generated from GPC are usually far from Gaussian shape, so I'm not sure any deconvolution would help in this type of analysis.
Please, could you send us the *.prm file or two with the mentioned peak cluster to support@dataapex.com for some testing?
Daniel Mentlí­k
DataApex

mkawatasu
Posts: 4
Joined: Wed Jun 03, 2009 11:07 am

Re: Gaussian peak fitting

Post by mkawatasu »

Dear Daniel,
Thank you for your reply.
I was conscious of several inflection point on the first peak. But this peak is one compound. In this case, ELSD is used as a detector, so peak shape is sometimes distorted. Actually, the peak is one peak.
I think the user would like to take Gaussian peaks simply created corresponding to "Local maximum" of chromatographic curve recognized as sample peaks.
I will send the prm file to you by e-mail.
Miyako Kawakatsu

Post Reply