Using partial least squares and principal component regression in simultaneous spectrophotometric analysis of pyrimidine bases

• Main Authors: Habibollah Khajehsharifi, Zahra Eskandari, Neda Sareban
• Format: Article
• Language: English
• Published: Elsevier 2017-02-01
• Series: Arabian Journal of Chemistry
• Subjects: Uracil; Cytosine; Thymine; Spectrophotometry; Partial least squares; Principal component regression
• Online Access: http://www.sciencedirect.com/science/article/pii/S1878535212001748

This study, for the first time, applies multivariate spectrophotometric calibration for simultaneous determination of three pyrimidine bases including uracil (URA), cytosine (CYT) and thymine (THY). Although determination of these bases is of great importance from a physiological and pharmaceutical perspective, it could be a difficult task since there is some sort of spectral overlapping. The principal component regression (PCR) and partial least-squares (PLS) model were used to overcome this problem and to construct the calibration sets containing URA, CYT and THY in the concentration range of 1.12–22.42, 1.11–27.78 and 1.26–25.22 μg mL−1, respectively. The absorption spectra were recorded from 220–320 nm. The results showed that the NPCs for URA, CYT and THY were 6, 5 and 4 by PCR and 4, 5 and 3 by PLS, respectively. In addition, the RMSEPs for URA, CYT and THY were 0.7067, 0.5093 and 0.6371 by PCR and 0.5469, 0.2700 and 0.5087 by PLS, respectively. The proposed method yielded recoveries ranging from 93.85 to 107.45 by PLS and 90.48 to 111.42 by PCR. The method was successful in simultaneous determination of the three bases in urine, serum and plasma.