Recursion workflow problems

[Cynthia]

Hi all,

I've been using the new recursion workflow algorithm from agilent, but it seems not to be working well.. I ran urine samples (minimum preparatory steps involved - metabolomics method) on the LC-TOF-MS and did the MFE extraction and exported the .cef files. I then created a new experiment on MassProfiler and added all the cef files from the samples to create an ion list (like a library).. well, it gave me 17 000 compounds in the (aligned) list!

Very well, I imported the ion list in masshunter in the find by formula section as a cef file and now I am running the find by formula algorithm on my data files using the cef file as reference list, but it is taking me almost 4 hours per sample! (And I have 47 samples to run)

Is there another quicker (but still accurate) way to run the find by formula algorithm to align my data? I have a report to write and don't have much time to do it.

Please help if you can, I'd greatly appreciate it.

Thank you all.

[lmh]

Am I right in thinking that you are doing metabolomics but are aiming to find empirical formulae for all possible ions?

[Cynthia]

Well, at this stage I just want `n reliable data matrix with my samples and compounds which I can trust not to have false positives and false negatives.

I am not currently worried about identification. I am going to do unsupervised statistics on the matrix, so metabolite names is not an issue right now.

[lmh]

OK, I thought for a moment you were trying to get empirical formulae for all your metabolomic data before doing the unsupervised statistics.

At this point I must come clean and admit I haven't done this with MassHunter. I currently use xcms for feature-finding (i.e. reducing data-set to a list of intensities corresponding to known, aligned RT/mass pairs. I can't say how it would perform on your data-set, but 4 hours seems extreme. It works on mzXML files, which I believe MassHunter should be able to export. It can also use .cdf, but that's not so good.