Cite this article as:

Chernova R. K., Monakhova Y. B., Varygina O. V. Chemometric Method of PLS in the Treatment of Titrimetric Data when Opredeleniye of Arginine and Lysine in Mixed Solutions. Izvestiya of Saratov University. New series. Series: Chemistry. Biology. Ecology, 2017, vol. 17, iss. 3, pp. 280-285. DOI: https://doi.org/10.18500/1816-9775-2017-17-3-280-285


Heading: 
UDC: 
543. 25
Language: 
Russian

Chemometric Method of PLS in the Treatment of Titrimetric Data when Opredeleniye of Arginine and Lysine in Mixed Solutions

Abstract

Arginine and lysine are the most important α-amino acids that are part of many proteins, which play an important role in the processes of nitrogen metabolism, growth and restoration of body tissues. It is used in medical nutrition in the postoperative recovery period, as well as as diagnostic factors. The simple express methods of separately determining the main α-amino acids of arginine and lysine in their mixed solutions are in demand. In this paper, the possibility of using the chemometric projection method for latent structures to determine lysine and arginine in binary mixtures from the pH-metric titration curves has been studied. 24 binary lysine-arginine mixtures with a titrimetric data matrix for chemometric analysis of 24 × 2682 were investigated. The second data block was the molar concentrations of arginine and lysine, 2 × 24, respectively. An alternative approach was used in the work – multidimensional calibration was constructed by the method of FL. The PL models were constructed for each of the amino acids separately. The root-mean-square error of the RMSEC calibration and the correlation coefficient R2 for arginine were RMSEC = 4 × 10-4M, R2 = 0.98, for lysine R2 = 0.98, RMSEC 4 × 10-4M. The test was carried out by the test-validation method. All samples are randomly divided into a training set (19 samples) and a test (5 samples) data sets. The breakdown into the training and test kits was randomly performed ten times. Training sets of samples were used to construct the PLS models with the help of which the analysis of “new” mixtures was carried out. A full cross-validation is applied for reduced training sets to test the optimal number of significant latent variables. In this case, the number of latent variables (PLS factors) in reduced data sets coincided with their number for complete models, which indicates their representativeness. The optimal number of factor PLs corresponded to seven for complete and training data sets for both analytes. The mean square prediction error (RMSEP) was 9.1 × 10-4 M and 9.8 × 10-4 M for arginine and lysine, respectively. It is shown that the chemometric method of PLS can be used for the separate determination of lysine and arginine in the joint presence on the basis of modeling the titration curves of their mixtures. Quantitative analysis of real objects requires the use of calibration mixtures containing a greater number of components.

References

1. Чернова Р. К., Варыгина О. В., Березкина Н. С. Избирательное определение гистидина в смешанных растворах α-аминокислот // Изв. Сарат. ун-та. Нов. сер. Сер. Химия. Биология. Экология. 2015. Т. 15, вып. 4. С. 10–15.

2. Monakhova Yи. В., Astakhov S. A., Kraskov A. V., Mushtakova S. P. Independent components in spectroscopic analysis of complex mixtures // Chemom. Intell. Lab. Syst. 2010. Vol. 103, № 3. P. 108–115.

3. Монахова Ю. Б., Астахов С. А., Муштакова С. П., Грибов Л. A. Методы декомпозиции спектров различной природы в анализе смесей сложного состава // Журн. аналит. химии. 2011. Т. 66, № 4. С. 361–372.

4. Монахова Ю. Б., Цикин А. М., Муштакова С.П. Метод независимых компонент как альтернатива методу главных компонент и дискриминантным алгоритмам в обработке спектрометрических данных // Журн. аналит. химии. 2015. Т. 70, № 9. С. 925–932.

5. Монахова Ю. Б., Кубалпа Т., Лахенмайер Д. В. Xемометрические методы в ЯМР-спектроскопическом анализе пищевых продуктов // Журн. аналит. химии. 2013. T. 68, № 9. С. 837–849.

6. Moisio Т., Heikonen М. A simple method for the titration of multicomponent acid-base mixtures // Anal. Bioanal. Chem. 1996. Jan. Vol. 354(3). P. 271–278.

7. Liao L., Yang J., Yuan J. Process monitored spectrophotometric titration coupled with chemometries for simultaneous determination of mixtures of weak acids // Anal. Chim. Acta. 2007. Vol. 591 (1). P. 123–154.

8. Rio V. del, Larrechi M. S., Callao M. P. Sequential injection titration method using second-order siarals: determination of acidity in plant oils and biodiesel samples // Talanta. 2010. Vol. 81. P. 1572–1579.

9. Rasouli Z., Ghavami R. Investigating the discrimination potential of linear and nonlinear spectral multivariate calibrations for analysis of phenolic compounds in their binary and ternary mixtures and calculation pKa values // Spectrochim Acta A Mol Biomol Spectrosc. 2016. Vol. 165. P. 191– 200.

Short text (in English): 
Full text (in Russian):