TY - JOUR
T1 - Quality Control of Targeted Plasma Lipids in a Large-Scale Cohort Study Using Liquid Chromatography–Tandem Mass Spectrometry
AU - Hirayama, Akiyoshi
AU - Ishikawa, Takamasa
AU - Takahashi, Haruka
AU - Yamanaka, Sanae
AU - Ikeda, Satsuki
AU - Hirata, Aya
AU - Harada, Sei
AU - Sugimoto, Masahiro
AU - Soga, Tomoyoshi
AU - Tomita, Masaru
AU - Takebayashi, Toru
N1 - Funding Information:
This research was funded by the Advanced Genome Research and Bioinformatics Study to Facilitate Medical Innovation (GRIFIN) from the AMED (JP18km0405209 for A.H. (Akiyoshi Hirayama)), JSPS KAKENHI (JP20H05743 for M.S.), the JST OPERA Program (JPMJOP1842 for S.H., M.S., and T.T.), and by grants from the Yamagata prefectural government and the city of Tsuruoka.
Publisher Copyright:
© 2023 by the authors.
PY - 2023/4
Y1 - 2023/4
N2 - High-throughput metabolomics has enabled the development of large-scale cohort studies. Long-term studies require multiple batch-based measurements, which require sophisticated quality control (QC) to eliminate unexpected bias to obtain biologically meaningful quantified metabolomic profiles. Liquid chromatography–mass spectrometry was used to analyze 10,833 samples in 279 batch measurements. The quantified profile included 147 lipids including acylcarnitine, fatty acids, glucosylceramide, lactosylceramide, lysophosphatidic acid, and progesterone. Each batch included 40 samples, and 5 QC samples were measured for 10 samples of each. The quantified data from the QC samples were used to normalize the quantified profiles of the sample data. The intra- and inter-batch median coefficients of variation (CV) among the 147 lipids were 44.3% and 20.8%, respectively. After normalization, the CV values decreased by 42.0% and 14.7%, respectively. The effect of this normalization on the subsequent analyses was also evaluated. The demonstrated analyses will contribute to obtaining unbiased, quantified data for large-scale metabolomics.
AB - High-throughput metabolomics has enabled the development of large-scale cohort studies. Long-term studies require multiple batch-based measurements, which require sophisticated quality control (QC) to eliminate unexpected bias to obtain biologically meaningful quantified metabolomic profiles. Liquid chromatography–mass spectrometry was used to analyze 10,833 samples in 279 batch measurements. The quantified profile included 147 lipids including acylcarnitine, fatty acids, glucosylceramide, lactosylceramide, lysophosphatidic acid, and progesterone. Each batch included 40 samples, and 5 QC samples were measured for 10 samples of each. The quantified data from the QC samples were used to normalize the quantified profiles of the sample data. The intra- and inter-batch median coefficients of variation (CV) among the 147 lipids were 44.3% and 20.8%, respectively. After normalization, the CV values decreased by 42.0% and 14.7%, respectively. The effect of this normalization on the subsequent analyses was also evaluated. The demonstrated analyses will contribute to obtaining unbiased, quantified data for large-scale metabolomics.
KW - cohort study
KW - lipid
KW - liquid chromatography–mass spectrometry
KW - quality control
KW - targeted lipidomics
UR - http://www.scopus.com/inward/record.url?scp=85153708606&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85153708606&partnerID=8YFLogxK
U2 - 10.3390/metabo13040558
DO - 10.3390/metabo13040558
M3 - Article
AN - SCOPUS:85153708606
SN - 2218-1989
VL - 13
JO - Metabolites
JF - Metabolites
IS - 4
M1 - 558
ER -