Page 1 of 1

The best method

Posted: Tue Jun 11, 2013 3:59 pm
by smartin
For a LC/MS-MS method, some possibilities:

METHOD 1: ISOCRATIC
Variation of the retention time: 0.02min
CV (between samples): around 3.70%
Width of the peak: 0.43
Total run time: 8 min

METHOD 2: GRADIENT
Variation of the retention time: 0.02min
CV (between samples): around 5%
Width of the peak: 0.23
Total run time: 8 min

METHOD 3: GRADIENT (only addinig 3 minuts of equilibration to the previous one)
Variation of the retention time: 0.0min
CV (between samples): around 6%
Width of the peak: 0.23
Total run time: 11 min

Only with this data.... which one do you choose?
I know it is a great discussion, but it will be nice a good discussion about it.

Thank you!!!!

Re: The best method

Posted: Wed Jun 12, 2013 2:28 am
by DR
It depends on myriad as yet unspecified factors.

Re: The best method

Posted: Wed Jun 12, 2013 2:53 am
by Don_Hilton
It all boils down to cost/benefit. What criteria are you trying to meet. Depending on what you are doing, cycle time can be the critical factor -- or LOD can be the critical factor. Or, dollar cost per analysis - which includes manpower for instrument repairs and troubleshooting. (And the differences in CV you gave - without a value of n doesn't mean a whole lot. For CV's varying by a factor of less than two, you have to have many replicates before an F test will tell you that they actually differ.)

Are we looking to do 20 samples for a whole study or 180/day per instrument across years?

I am reminded of being asked which pair of earrings looked better. I picked one - which resulted in the dress selected for the evening no longer being suitable. So, while she went back through the closet.... (Do not make a selection unless you have sufficient information.)

Re: The best method

Posted: Wed Jun 12, 2013 7:24 am
by smartin
Thanks,

I know that the information is not enough. I'm agree on it.

Regarding some of the questions, this method is mainly thought for a study, a little one.
Maybe it could be implemented for the routine analysis. but I'm sure that it should be modified, as it could incorporate some other related molecules (and analyse all of them together).
So, by now, for this study, the time required for the analysis is not critical.
But LOD is important!!!!

Regarding on the cost, what exactly means with it? Is a gradient method more expensive than an isocratic one?
Regarding the "variation of the retention time", how much is allowed? It means, considered acceptable?
Regarding the "width of the peak", how much is allowed? It means, considered acceptable?

Thank you a lot for your help, and also your recommendations.