Large-scale multi-component quantitative data correction method and system based on IS combined with SVR

Abstract

The invention provides a large-scale multi-component quantitative data correction method based on the combination of IS and SVR, and the method comprises the steps: 1) making QC samples, and carrying out the split charging and storage of the QC samples; (2) sequentially detecting the actual samples in batches, inserting the QC samples into the actual samples, and carrying out metabonomics analysis by a chromatography-mass spectrometry method to obtain biological metabonomics data; 3) using QC sample data as training data, establishing an IS scaling and support vector regression abundance prediction model, and predicting random system errors in the data; 4) substituting the metabolite peak area parameter of the actual sample data into the prediction model, removing the random system error in the actual sample data, and obtaining the normalized metabolite peak area of each actual sample s; and 5) fitting a linear calibration curve between the corrected metabolite peak area and the actual metabolite concentration, and calculating the metabolite concentration value. According to the method, system errors in the data are comprehensively removed, the data normalization efficiency is improved, and the biological repeatability and the result accuracy of quantitative data are improved.

Description

technical field [0001] The invention belongs to the technical field of biological metabolomics, and in particular relates to a large-scale multi-component quantitative data correction method and system based on IS combined with SVR. Background technique [0002] Metabolomics is an emerging frontier technology that has emerged in the fields of biomedicine, food safety and environmental toxicology in recent years. Due to the diversity of biological states and lifestyles of individual organisms, metabolomic studies of large cohort samples have become more and more widely used due to their advantages in terms of average biological heterogeneity; high-throughput LC-MS technology The development of large-scale metabolomics has also further promoted the development of large-scale metabolomics. During large-scale, long-term metabolomics experiments, derived from pre-analytical stages (sample collection, sample storage conditions and time, sample preparation, different operators, et...

AI Technical Summary

Problems solved by technology

Therefore, the normalization method based on QC cannot simulate the influence of matrix effect well

Images