A Software\nTool for Rapid and Automated Preprocessing\nof Large-Scale Serum Metabolomic Data by Multisegment Injection-Capillary\nElectrophoresis-Mass Spectrometry

Mass spectrometry (MS)-based metabolomics often rely\non separation\ntechniques when analyzing complex biological specimens to improve\nmethod resolution, metabolome coverage, quantitative performance,\nand/or unknown identification. However, low sample throughput and\ncomplicated data preprocessing procedures remain major barriers to\naffordable metabolomic studies that are scalable to large populations.\nHerein, we introduce PeakMeister as a new software tool in the R statistical\nenvironment to enable standardized processing of serum metabolomic\ndata acquired by multisegment injection-capillary electrophoresis-mass\nspectrometry (MSI-CE-MS), a high-throughput separation platform (<4\nmin/sample) which takes advantage of a serial injection format of\n13 samples within a single analytical run. We performed a rigorous\nvalidation of PeakMeister by analyzing 47 cationic metabolites consistently\nmeasured in 5,000 serum and 420 quality control samples from the Brazilian\nNational Survey on Child Nutrition (ENANI-2019) comprising a total\nof 224,983 metabolite peaks acquired in 40 days across three batches\nover an eight-month period. A migration time index using a panel of\n11 internal standards was introduced to correct for large variations\nin migration times, which allowed for reliable peak annotation, peak\nintegration, and sample position assignment for serum metabolites\nhaving two flanking internal standards or a single comigrating stable-isotope\ninternal standard. PeakMeister accelerated data preprocessing times\nby 30-fold compared to manual processing of MSI-CE-MS data by an experienced\nanalyst using vendor software, while also achieving excellent peak\nannotation fidelity (median accuracy >99.9%), acceptable intermediate\nprecision (median CV = 16.0%), consistent metabolite peak integration\n(mean bias = −2.1%), and good mutual agreement when quantifying\n16 plasma metabolites from NIST SRM-1950 (mean bias = −1.3%).\nReference ranges are also reported for 40 serum metabolites in a national\nnutritional survey of Brazilian children under 5 years of age from\nthe ENANI-2019 study. MSI-CE-MS in conjunction with PeakMeister allows\nfor rapid and automated processing of large-scale metabolomic studies\nthat tolerate nonlinear migration time shifts without complicated\ndynamic time warping or effective mobility scale transformations.