Hi,
I have the raw data (chemical shifts and intensities) obtained from NMR experiment in blood samples. How can we recognize the metabolites and concentration present per participant? We have the chemical shift (9.998...9.999) and intesities only.
Please send me any approach.
Thanks!!
There are some softwares and plataforms that you can use as Chenomix or Bayesil for NMR metabolites.
While for the interpreting the datas you should use some Statistical approaches as PCA PLSDA analysis or heatmaps. However first it is need to binning the NMR data for statistic analysis using NMR data.
Thank you! However we have no spectra but results as raw data contained in spreadsheets. Such programs recognize metabolites by analyzing spectra only.
Yes, spectra only. But you can transform one spectra again from the data in spreadsheet?! I think it is possible. I do not have sure, because I think you must have the original FID for this. But you can try. Or ask the original FIDs from the person who have performed the analysis.
Dear Claudia,
you may want to have a look on
http://www.hmdb.ca/spectra/nmr/one_d/search/new
- But without raw-data (multiplicities, lineshape) I'd guess it will be difficult to dig out your metabolites reliably!
- For absolute vale quantification (mg/ml) you need experimental data of the spectra (receiver gain, number of scans, and and and...). NMR Intensities do not reflect molcar concentatins. You need the integrals!
- If you are interested in relative concentrations only and if you can be sure that all your data have be acquired with identical experimental and processing parameters the intensities might do as relative measures.
I hope this helps - good luck
Alfred
Thank you, Alfred. You're right, HMDB is useful. However we have thousends of signals and more than 200 patients. Do you know if there is an authomatic way to do that?
Dear Claudia, have a look at:
http://metabolomicssociety.org/resources/metabolomics-software
...personally I do not have experience with any of these tools.
We have been using a in-house database of metabolite reference spectra together with Amix software (Bruker) and experienced NMR people...
As proposed by Banny...I would "boil down" your data by use of PCA/PLS/PLSDA to a subset of significant changing signals. With those you might go back to HMDB for identification.
P.S. I guess you have human data. Are you looking for changes of endogenic metabolites (part of HMDB) or drug/food/livestyle related compounds (mostly not part of HMDB) - and how do you want to differentiate both classes - which is a art by itself (my own experience!).
- Badges
- Science topic
- Similar topics
- Biological Science
- Molecular Biology
- Biomolecules
- Nucleic Acids
- RNA