Premium
Principal component analysis of molecularly based signals from infant formula contaminations using LC‐MS and NMR in foodomics
Author(s) -
Inoue Koichi,
Tanada Chihiro,
Hosoya Takahiro,
Yoshida Shuhei,
Akiba Takashi,
Min Jun Zhe,
Todoroki Kenichiro,
Yamano Yutaka,
Kumazawa Shigenori,
Toyo'oka Toshimasa
Publication year - 2016
Publication title -
journal of the science of food and agriculture
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.782
H-Index - 142
eISSN - 1097-0010
pISSN - 0022-5142
DOI - 10.1002/jsfa.7584
Subject(s) - principal component analysis , gas chromatography–mass spectrometry , chemistry , component (thermodynamics) , chromatography , mass spectrometry , computational biology , biology , computer science , artificial intelligence , physics , thermodynamics
BACKGROUND The challenge in developing analytical assessment of unexpected excess contaminations in infant formula has been the most significant project to address the widespread issue of food safety and security. Foodomics based on metabolomics techniques provides powerful tools for the detection of tampering cases with intentional contaminations. However, the safety and risk assessments of infant formula to reveal not only the targeted presence of toxic chemicals, but also molecular changes involving unexpected contaminations, have not been reported. In this study, a huge amount of raw molecularly based signals from infant formula was analysed using reversed phase and hydrophilic interaction chromatography with time‐of‐flight MS ( LC‐MS ) and 1 H nuclear magnetic resonance ( NMR ) and then processed by a principal component analysis ( PCA ). RESULTS PCA plots visualised signature trends in the complex signal‐data batches from each excess contamination of detectable chemicals by LC‐MS and NMR . These trends in the different batches from a portion of excess chemical contaminations such as pesticides, melamine and heavy metals and out‐of‐date products can be visualised from spectrally discriminated infant formula samples. CONCLUSION PCA plots provide possible attempts to maximise the covariance between the stable lot‐to‐lot uniformity and excess exogenous contaminations and/or degradation to discriminate against the molecularly based signals from infant formulas. © 2015 Society of Chemical Industry