Metabolomics Society Forum

Software => XCMS Online => Topic started by: MetGG on March 29, 2012, 03:04:27 AM

Title: process UHPLC-QTRAP data
Post by: MetGG on March 29, 2012, 03:04:27 AM
Hi,

I wish to acquire UHPLC-QTRAP data in scan mode and use XCMS online to process them. I have selected HPLC-single quad as a starting point, but I need to adapt some parameters since UHPLC is to be used (10min gradient). From the pre-defined parameters, I am tempted to change:

1. Feature detection>FWHM, 3-6 sec
2. Feature detection>Step at 0.1or a bit more?
3. Retention time correction>Initial grouping:bw, 3-6?
4. Alignment>bw, 2-4 sec?

All the other parameters would remain unchanged. Does it look acceptable?

And a last question: can XCMS online deal with both centroid and profile data or one is better than the other? I guess if a step window of 0.1 Da is used, centroid would be better?

Thanks for your help,
Best regards
Title: Re: process UHPLC-QTRAP data
Post by: Ralf on March 29, 2012, 10:37:45 AM
Quote
2. Feature detection>Step at 0.1or a bit more?
why would you want increase the step size for high res data ?
The other parameters look OK.

If you use centroid mode data, try using centWave for feature detection.
See other posting for more details on that.
Title: Re: process UHPLC-QTRAP data
Post by: MetGG on March 29, 2012, 01:01:18 PM
Sorry I omitted to mention it: the QTRAP operates at low resolution, like a triple quadrupole. That's why I intend to use the template for HPLC-single quad, but adapt a bit parameters to take into account the fact that UPLC is employed. So I thought maybe an increased step size will perform better.

Could I still use Centwave on low resolution data?

Thanks for you answer
Title: Re: process UHPLC-QTRAP data
Post by: Ralf on March 30, 2012, 03:34:15 PM
OK, then it makes sense.

Yes, you can still use centWave for low resolution data, you would just have use a pretty high value for the ppm parameter.
Title: Re: process UHPLC-QTRAP data
Post by: MetGG on April 02, 2012, 08:16:42 AM
Thanks for your reply.