| Modern mass spectrometers produce a large amount of mass spectrum data(mass spectrum fingerprint),which were usually used to identify the molecular species and composition of the whole or components of pure substances or complex systems,which requires specialized analysis and time consuming.It is undoubtedly an innovative methodological research to directly distinguish authenticity of complex ingredients and trace the products by using mass spectrum fingerprints.This research aims to explore the mass spectrum fingerprint of six livestock milk,and to establish the authenticity identification model based on milk fat mass spectrum fingerprint.Collecting 56 representative samples of Mongolian horse,Holstein cow(cow),Bactrian camel,goat,yak and buffalo milk were collected,milk fat extracted by n-hexane and screened by supercritical fluid chromatograph quadrupole time of fight mass spectrometry(SFC-Q-TOF-MS)with electrospray ionization(ESI~+).TAGs in six livestock milk fat were identified for PCA analysis and used as standard comparison to evaluate the clustering effect of mass spectrum fingerprint.The effective retention period,specific mass charge ratio and molecular ion relative abundance of BPI and MS images were intercepted in three dimensions,and multiple groups of mass spectrum fingerprint data were obtained.Principal component analysis(PCA)and hierarchical cluster analysis(HCA)were performed to observe the characteristics and clustering of mass spectrum fingerprint of six livestock milk.At the same time,TAGs molecules were identified and PCA was carried out to compare and evaluate the clustering effect of PCA by mass spectrum fingerprint,and the best clustering effect was selected to establish SIMCA model.BPI showed that the peak time of each species and the overall chromatographic peak morphology have obvious species differences,which suggested that it is feasible to establish the identification of species authenticity by chromatographic fingerprint.113 TAGs were identify from six livestock milk from n-hexane extracts with molecular weights of 470～880.PCA analysis of milk samples was based on species clustering and the results were taken as the standard.In the mean time it was conducted directly on the mass spectrum to observe the clustering effect of each mass spectrum segment on milk samples.BPI were intersected with retention time of5-14min.PCA with total mass spectrum fingerprint data of six livestock milk and reference to the mass spectrum range of molecular ions of TAGs in milk was performed with the best results obtained by intercepting 480-910 m/z on the horizontal axis and 30%relative abundance on the vertical axis,mass spectrum fingerprint of 3092 variables(molecular ions and fragments)was obtained,and the results were consistent with the analysis results of specific components and content of TAGs.The clustering characteristics of mass spectrum fingerprint conform to species taxonomy.As the results,milk samples clustered in to six groups,and distances among groups were consistent with their species taxonomy.Milk sample of horse and Bactrian camel and that of four bovine family animal were clustered separately in to three phase of 3D space;among bovine family,cow,yak,buffalo and goat milk samples grouping closely,while the clustering distance of Bos(cow and yak),Bubalus(buffalo)and Capra(goat)were consistent to the natural classification of these Bovidae animals.Its was also found that the organic and non-organic Holstein milk were clustering separately,and indicating TAGs are different from FA fingerprints.HCA was consistent with PCA results.Establishing SIMCA discrimination model based on training sample set composed of six livestock milk mass spectrum fingerprints,the accuracy of internal and external verification is 100%and 95%respectively.It was feasible to use milk fat mass spectrum fingerprint to quickly identify the authenticity,which provided an innovative idea,strategy and technology for the authenticity identification and traceability of livestock milk. |
PDF Full Text Request