A chemometrics-based approach for the determination of thymoquinone from Nigella sativa L. (Black Cumin) seeds of different geographical regions using the HPLC technique

Abstract

Background and Aims: In this study, thymoquinone (TQ) from black cumin will be quantified from several geographical regions, including India, Syria, Saudi Arabia, Iraq, and Turkey. Additionally, to forecast the chromatographic behavior of the analyte in artificial intelligence (AI)-based models, the study used both ensemble machine learning methodologies and chemometrics-based approaches. Methods: An Agilent Technologies (1200 series, USA) instrument that includes an autosampler, a binary pump, a diode array detector (DAD), and a vacuum degasser was used for the HPLC analysis. Using five different single models—principal component regression (PCR), least square-support vector machine (LSSVM), least square boost (LSQ-BOOST), adaptive neuro-fuzzy inference system (ANFIS), and step-wise linear regression—the HPLC-DAD technique was used to simulate the qualitative and quantitative properties of TQ (SWLR). Results: The collected results demonstrated that samples from India and Iraq have the highest concentration of TQ. TQ was present in all samples, but in varying amounts; the amounts of TQ in the samples from Iraq, India, Saudi Arabia, Syria, and Turkey, respectively, were 0.031, 0.030, 0.022, 0.005, and 0.001%. According to a comparison of their performances, the four ensemble machine learning techniques can reproduce the chromatographic properties of TQ, PA, and tR with minimum and maximum NSE-values of 0.842 and 0.999 in the training phase and 0.918 and 0.999 in the testing phase, respectively. Conclusion: The TQ content of each sample of black cumin, which was collected from various geographical locations, was determined quantitatively. The quantity of thymoquinone fluctuates depending on geographic variances, according to HPLC data. Five different AI-based models, including SWLR, PCR, LSSVM, ANFIS, and LSQ-Boost, were used to simulate the chromatographic behavior of TQ information of retention duration and peak area using various independent factors. Additionally, SAE, WAE, NNE, and ANFIS-E are informed by the application of ensemble machine learning to enhance the performance of AI-based models. Comparing the approaches to the individual models, they both demonstrated lower error values in terms of RMSE and MSE.

Keywords

• Chemometrics
• HPLC
• Thymoquinone
• Black cumin
• geographical regions

References

1. Abba, S. I., Pham, Q. B., Usman, A. G., Linh, N. T. T., Aliyu, D. S., Nguyen, Q., & Bach, Q. V. (2020). Emerging evolu-tionary algorithm integrated with kernel principal component analysis for modeling the performance of a water treatment plant. Journal of Water Process Engineering, 33, 101081. doi: 10.1016/j.jwpe.2019.101081 google scholar
2. Chen, M., Wen, F., Zhang, Y., Li, P., Zheng, N., & Wang, J. (2019). De-termination of native lactoferrin in milk by HPLC on HiTrapTM Heparin HP column. Food Analytical Methods, 12(11), 25182526. doi: 10.1007/s12161-019-01572-x google scholar
3. Golmohammadi, H., Dashtbozorgi, Z., & Vander Heyden, Y. (2015). Support Vector Regression Based QSPR for the Prediction of Re-tention Time of Peptides in Reversed-Phase Liquid Chromatogra-phy. Chromatographia, 78(1-2), 7-19. doi: 10.1007/s10337-014-2819-1 google scholar
4. Herlina, Aziz, S. A., Kurniawati, A., & Faridah, D. N. (2017). Changes of Thymoquinone, Thymol, and Malondialdehyde Content of Black Cumin (Nigella sativa L.) in Response to Indonesia Trop-ical Altitude Variation. HAYATI Journal of Biosciences, 24(3), 156-161. doi: 10.1016/j.hjb.2017.08.004 google scholar
5. Isik, S., Kartal, M., & Erdem, S.A. (2017). Quantitative analysis of thy-moquinone in Nigella sativa L. (Black Cumin) seeds and commer-cial seed oils and seed oil capsules from Turkey. Ankara Univer-sitesi Eczacilik Fakultesi Dergisi, 41(1), 34-41. doi: 10.1501/Ecz-fak_0000000593 google scholar
6. Liu, Y., Kohlberger, T., Norouzi, M., Dahl, G. E., Smith, J. L., Mohtashamian, A., Stumpe, M. C. (2019). Artificial intelli-gence-based breast cancer nodal metastasis detection insights into the black box for pathologists. Archives of Pathology and Labora-tory Medicine, 143(7), 859-868. doi: 10.5858/arpa.2018-0147-OA google scholar
7. Marrero-Ponce, Y., Barigye, S. J., Jorge-Rodriguez, M. E., & Tran-Thi-Thu, T. (2018). QSRR prediction of gas chromatography retention indices of essential oil components. Chemical Papers, 72(1), 5769. doi: 10.1007/s11696-017-0257-x google scholar
8. Nourani, V., Hakimzadeh, H., & Amini, A. B. (2012). Implementation of artificial neural network technique in the simulation of dam breach hydrograph. Journal of Hydroinformatics, 14(2), 478. doi: 10.2166/hydro.2011.114 google scholar

9. Rezai, F., Işık, S., Kartal, M., & Aslan Erdem, S. (2018). Ef-fect of priming on thymoquinone content and in vitro plant re-generation with tissue culture of black cumin (Nigella sativa L.) seeds. Journal of Chemical Metrology, 12(2), 89-98. doi: 10.25135/jcm.18.18.09.950 google scholar
10. Shadrin, D., Pukalchik, M., Kovaleva, E., & Fedorov, M. (2020). Artifi-cial intelligence models to predict acute phytotoxicity in petroleum contaminated soils. Ecotoxicology and Environmental Safety, 194(February), 110410. doi: 10.1016/j.ecoenv.2020.110410 google scholar
11. Tham, S. Y., & Agatonovic-Kustrin, S. (2002). Application of the artificial neural network in quantitative structure-gradient elution retention relationship of phenylthiocarbamyl amino acids deriva-tives. Journal of Pharmaceutical and Biomedical Analysis, 28(3-4), 581-590. doi: 10.1016/S0731-7085(01)00690-2 google scholar
12. Usman, A. G., Işik, S., & Abba, S. I. (2021). Hybrid data-intelligence algorithms for the simulation of thymoquinone in HPLC method development. Journal of the Iranian Chemical Society, 18(7), 1537-1549. doi: 10.1007/s13738-020-02124-5 google scholar
13. Usman, Abdullahi Garba, Isik, S., Abba, S. I., & Mericli, F. (2021a). Artificial intelligence based models for the qualitative and quan-titative prediction of a phytochemical compound using HPLC method. Turkish Journal of Chemistry, 44(5). doi: 10.3906/kim-2003-6 google scholar
14. Usman, Abdullahi Garba, Işik, S., Abba, S. I., & Meriçli, F. (2021b). Chemometrics-based models hyphenated with ensemble ma-chine learning for retention time simulation of isoquercitrin in Coriander sativum L. using high-performance liquid chro-matography. Journal of Separation Science, 44(4), 843-849. doi: 10.1002/jssc.202000890 google scholar
15. Vasiljevic, T., Onjia, A., Cokesa, D., & Lausevic, M. (2004). Opti- google scholar
16. mization of artificial neural network for retention modeling in high-performance liquid chromatography. Talanta, 64(3), 785790. doi: 10.1016/j.talanta.2004.03.032 google scholar
17. Zribi, I., Omezzine, F., & Haouala, R. (2014). Variation in phyto-chemical constituents and allelopathic potential of Nigella sativa with developmental stages. South African Journal of Botany, 94, 255-262. doi: 10.1016/j.sajb.2014.07.009 google scholar