In metabolomic studies, what is the physiological normalization method indicated for human urine samples in aerobic exercise protocols?
Emwas et al. 20018 suggested normalization of specific gravity above creatinine.
Vollmar et al. 2019 evaluated that only specific gravity or PQN would be the most appropriate normalizations.
Creatinine was usually used in most studies.
However, in my tests, I consider that creatinine metabolism changes during exercise, and therefore, it would not be more prudent to use creatinine to normalize.
What do you think about this?
Mr. Salgueiro,
Quantification of complex mixture such as urine by NMR is unreliable due to the high LODs of NMR and error contribution from the fitting of the NMR patterns of the proton spectra.
Conversely, reliable quantification of urine is carried out by means of ESI-MS using the diffusion parameters according to our model formulas derived more recently. The quantification is carried out, directly, in solution per span pf the scan time. For analysis of steroids in mixture we have achieved reliability r = 0.99995 at pg.(mL)-1 concentration range. Such method performances are far of the capability of NMR.
Please consider detail in the following references together with the literature cited therein:
1. Book QUANTITATIVE RELATIONS AMONG TEMPERATURE, ANALYTE CONCENTRAT...
2. https://zenodo.org/record/3606820#.XoIUVfTgqUk
3. Article A mass spectrometric stochastic dynamic diffusion approach t...
reference 3, in particular, shows detail on the calculation task according to our theory and formulas together with quantification of cyclodextrins, where the analysis of b-CD has resulted to reliability r = 1.
As you suggest, urinary creatinine is affected by exercise. If you dont have total volume excreted or osmolality for each sample then PQN is a good option.
A broad consideration of the impact of scaling and normalisation on urine NMR data can be found in Craig A, et al Scaling and normalization effects in NMR spectroscopic metabonomic data sets. Anal Chem. 2006;78(7):2262-7. PubMed PMID: 16579606.
and a more focussed approach comparing normalisation effects in an exercise study in Daskalaki E, et al A study ofthe effects of exercise on the urinary metabolome using normalisation to individual metabolic output. Metabolites. 2015 ;5(1):119-39.
In reality comparing a couple of methods is always good practice
Dear Diego,
some time back we have published this little paper:
Article Probabilistic Quotient Normalization as Robust Method to Acc...
maybe it helps. I'd like to mention that it was cited, since in was published, in about 1000 references, so it has found some applications...
Kind regards
Alfred
PQN is a very good metho, despite the problem of requiring a reference.
A new proposition which is rather good is :
VSN: Variable sorting for normalization
Gilles Rabatel Federico Marini Beata Walczak Jean‐Michel RogerFirst published:29 July 2019
https://doi.org/10.1002/cem.3164
When we are preparing biological samples, we may not get the desired amount of extract/fluid for each sample. Hence we need to dilute the samples for arrive at the necessary volume of analysis for NMR analysis. Therefore, in order to compare the data by statistical means viz PCA we have to account for these different dilutions and bring them to the same scale. This preprocessing method is called normalization. From the practical point of view, during bulk sample preparation we may make mistakes in dilution of the precious sample(s). If these mistakes can be correctly tracked and identified we can account for these erroneous dilutions to the NMR data of these samples by performing normalization. Normalization takes into account the different sample dilutions and scales the spectra accordingly so that we can arrive at the same virtual overall concentration, that aids in comparison of data by statistical means.
- Badges
- Science topic
- Similar topics
- Methodology
- Research Methodology
- Research Papers