Premium
Estimation of Oil Content and Fatty Acid Composition in Cottonseed Kernel Powder Using Near Infrared Reflectance Spectroscopy
Author(s) -
Quampah Alfred,
Huang Zhuang Rong,
Wu Jian Guo,
Liu Hai Ying,
Li Jin Rong,
Zhu Shui Jin,
Shi Chun Hai
Publication year - 2012
Publication title -
journal of the american oil chemists' society
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.512
H-Index - 117
eISSN - 1558-9331
pISSN - 0003-021X
DOI - 10.1007/s11746-011-1945-2
Subject(s) - near infrared reflectance spectroscopy , partial least squares regression , chemistry , stearic acid , oleic acid , linoleic acid , fatty acid , palmitic acid , chromatography , analytical chemistry (journal) , near infrared spectroscopy , mathematics , organic chemistry , biochemistry , biology , statistics , neuroscience
Oil content and fatty acid composition in 444 ground cottonseed kernel samples were analyzed using near infrared reflectance spectroscopy (NIRS). Calibration equations were developed for oil and fatty acid contents with the modified partial least squares (MPLS) regression method. The correlations between NIRS and reference values in external validation were in agreement with the predictions in calibration. Each equation was assessed based on the relative prediction determinant for external validation (RPD v ). Equations corresponding to total oil content (RPD v = 11.495) and linoleic acid (RPD v = 5.026) showed high accuracy. For palmitic acid (RPD v = 1.914), myristic acid (RPD v = 1.724) and oleic acid (RPD v = 1.999), the equations were predicted with relatively high accuracy while those for palmitoleic acid (RPD v = 0.686), stearic acid (RPD v = 0.792), linolenic acid (RPD v = 0.475) and 1‐eicosenoic acid (RPD v = 0.619) were poorly predicted. The equations for traits with RPD v > 1.5 could be reliably used in screening samples for breeding programs.