Chromatography Forum

Hello,

I am looking for a solution for a co-worker and am not that familiar with the software he uses. He is running a Bruker Scion GCMS with the MS Workstation software. We are looking for a way to perform a NIST library search for all of the peaks in a chromatogram (very messy chromatograms). The software will allow you to search the library for an individual spectrum but not for every peak all at the same time. We need to process multiple samples and processing every spectrum is going to be take a prohibitively long time.

Does anyone know if you can perform a library search for every peak in a more automated way?

Any help would be greatly appreciated.

Thanks, Adam

This isn't the answer you want, but I'm writing it because I had someone recently ask for the same thing.

Even if there is a technical solution out there that will do this, you almost certainly do not want it! The problem is that every library search on every peak will produce multiple lines of results (there will be multiple hits in the NIST library). Some will be good, some will be certain rubbish. At the very least, you will need to review how good the fit is - the best match will, in some cases, be utterly useless, in others it will be spot on, and in others it will be a close relative, but the third match will be what you're looking for. How can the automated system present the data in such a way that a human can see the top X hits and how good they are, without the result becoming a huge unwieldy morass? It's not a simple one-peak, one-result spreadsheet (at least, if it is presented that way, it's a danger to reality!).

You're absolutely right, the time taken to search each peak individually will be prohibitive, but the time taken to review each peak individually after making the search is far worse, and it's necessary.

I'm convinced that at the moment, there are only two reasonable requests a client can make of an analyst: (1) measure the following range of named chemicals, or (2) look for a particular identifiable needle in my haystack (an active fraction, a difference between two treatments), and identify what this needle is. Anyone who wants you to identify every unknown in a complex sample should expect that it will take a very long time and cost a fortune, and probably generate a data-set that is too large for them to review.

Opening/Viewing Bruker (.xms) GC/MS files in AMDIS viewtopic.php?f=3&t=25755

Hi Adam,

yes, you can do that with OpenChrom:
https://openchrom.net

But I agree with lmh that blindly trust the NIST results isn't what you really want. OpenChrom offers to automatically fetch NIST identfications and to assign them to each peak in the chromatogram. Thereafer, OpenChrom allows you to review each peak, show the mass spectrum, the search results and a molecule view that will be rendered on-the-fly. You can also delete results that are not matching or add a manual identification. At each time, you are able to save the chromatogram, it's peaks and identifications in *.ocb format. So that you're able to proceed with your work at a later point.

We'll create a howto on our wiki soon.

Until then, please don't hesitate to contact us:
info [at] openchrom.net

Thank you all for your help.

I took a look at the Spectrum matching on the Bruker MSWS and it IS definitely giving some sketchy matches for many peaks. I'm just going to have to talk with my Lab Manager about finding a solution that will provide some quality useful data for the client. This was my first post on this forum and I am impressed by the speed and quality of responses received. Hopeful I can help someone else out here in the future.

Thanks, Adam

Hi Adam,

as a follow-up to the post of eselmeister here are links to the Wiki describing how to perform peak detection and identification with OpenChrom and how to set up and run a batchwise processing of datasets:

https://wiki.openchrom.net/index.php/De ... tify_peaks
https://wiki.openchrom.net/index.php/Batch_Processing

Cheers,
AK