A new method for discrimination and quantification of true biological signals in LC-MS-based metabolomics analysis
Figure. The strategy and evaluation of LC-MS-based metabolomics analysis. (A) The design of the metabolomics experiment. (B) Overlay of LC-MS total ion chromatograms for a dilution series of a sample and the blank samples to discriminate false positives. (C) Benchmarking the five steps of peak filtering using artificial samples. (D) Principle components analysis (PCA) models for rice samples. Left, the traditional method; right, the new strategy. (E) Venn diagram of the number of differential peaks identified in the traditional (circle on the left) versus the new metabolomics approach (circle on the right). (F) and (G) The loading S-plots for PCA with the traditional metabolomics method (F) and the new strategy using RCI (G).
Metabolomics is one of the most important parts of system biology. It has been widely applied to the research of gene function and biomarkers. With the rapid development of high sensitive, high resolution mass spectrum, it becomes easy to detect thousands of metabolites. However, LC-MS unavoidably yields a large number of false positive signals mixed with true biological signals. Without the means to confidently discriminate and evaluate the detected signals, biomarker discovery turns out to be misleading, or downright impossible.
The research from Prof Qi’s group developed a new strategy to prepare a blank sample, a Quality Control mixture (QC_mix) that combines all samples in equal proportions, and a dilution series of the QC_mix. Subseqyently, a hierarchical five-step filtering approach to discriminate the biological and non-biological signals and evaluate the quantitative performance of each peak. This new strategy introduces a relative concentration index (RCI) as an arbitrary concentration to build a relative quantitative model for the calibration and normalization of each analyte. The results showed 92.4% of the peaks were eliminated as false positives or peaks with insufficient quantitative performance in the artificial samples and 71.4% false positive were eliminated from rice seeds samples .
This research is publishing on Molecular Plant. Dr. Lixin Duan is the first author of this paper. This work was supported by the Chinese National Key Programs for Research and Development, the Key Project of Chinese National Programs for Fundamental Research and Development, the National High-tech R&D Program of China and the National Natural Science Foundation of China.
Paper linkage: http://dx.doi.org/10.1016/j.molp.2016.05.009
Metabolomics is one of the most important parts of system biology. It has been widely applied to the research of gene function and biomarkers. With the rapid development of high sensitive, high resolution mass spectrum, it becomes easy to detect thousands of metabolites. However, LC-MS unavoidably yields a large number of false positive signals mixed with true biological signals. Without the means to confidently discriminate and evaluate the detected signals, biomarker discovery turns out to be misleading, or downright impossible.
The research from Prof Qi’s group developed a new strategy to prepare a blank sample, a Quality Control mixture (QC_mix) that combines all samples in equal proportions, and a dilution series of the QC_mix. Subseqyently, a hierarchical five-step filtering approach to discriminate the biological and non-biological signals and evaluate the quantitative performance of each peak. This new strategy introduces a relative concentration index (RCI) as an arbitrary concentration to build a relative quantitative model for the calibration and normalization of each analyte. The results showed 92.4% of the peaks were eliminated as false positives or peaks with insufficient quantitative performance in the artificial samples and 71.4% false positive were eliminated from rice seeds samples .
This research is publishing on Molecular Plant. Dr. Lixin Duan is the first author of this paper. This work was supported by the Chinese National Key Programs for Research and Development, the Key Project of Chinese National Programs for Fundamental Research and Development, the National High-tech R&D Program of China and the National Natural Science Foundation of China.
Paper linkage: http://dx.doi.org/10.1016/j.molp.2016.05.009