BSc in School of Medical Laboratory and Biotechnology,
Chung Shan Medical University
Current AS:MIT student.
Finger Printing Meats Using High Resolution Mass Spectrometry
Due to the concerns for food authenticity rising in recent years, many methods of classifying different types of meats have been designed using various techniques. The authenticity of meat species was performed in this project using high resolution mass spectrometry, to analyse trypsin digests of 5 different types of meats including pork, beef, lamb, chicken and turkey. With the aid of multivariate data analysis, the meats were successfully classified. The outliers were firstly picked out on a PCA model using Hostelling’s T-squared test, followed by an OPLS-DA model which was trained to separate the five kinds of meats. Differentiation was successfully achieved due to the better group separation ability of OPLS-DA, but the white meats were considered as outliers since they are out of 95% significant level. The final model was trained using meat cuts of beef, lamb and pork. The variables that strongly affect the separation were determined from the loading plot. These variables that are unique to a particular type of meat can be potential biomarkers that can benefit from further analysis of MS/MS. The relationship between processed meats and the model trained using meats cuts was determined using T-predicted scores plot. Some processed meat products were found to be cross-contaminated during the produce manufacture process or adulterated of other kinds of meat. The OPLS-DA model is further optimized by adjusting the number of most intensity peaks used to train the model. The study demonstrates a fast analysis method to classify three types of meats, which only requires MS scans of the analytes.
Other Academic Work
Naïve hESC : history, mechanisms, and applications