Chromatographic retention time alignment method based on primary spectrogram and deep learning

Abstract

The invention discloses a chromatographic retention time alignment method based on a primary spectrogram and deep learning. The method comprises the following steps: 1) extracting an ion flow chromatographic peak of each mass spectrum file in each sample; 2) randomly selecting one sample as a reference sample, and correcting the ion current chromatographic peak of the corresponding sample according to the average retention time deviation between the ion current chromatographic peak of other samples in each retention time window and the reference sample; 3) constructing feature vectors according to retention time information and intensity information of ion flow chromatographic peaks of the two corrected samples to be aligned; 4) labeling the feature vector according to an existing identification result; 5) training a deep learning classification model by using the marked feature vectors; and 6) for two pieces of mass spectrum data to be aligned, constructing feature vectors corresponding to the two pieces of mass spectrum data to be aligned, and inputting the feature vectors into the trained classification model for alignment. According to the method, an alignment result can be given according to primary spectrum information, and analysis of trace compounds is facilitated.

Description

technical field [0001] The invention relates to a method for aligning chromatographic retention times in analytical chemistry, in particular to a method for aligning chromatographic retention times between different samples in shotgun proteomics. Background technique [0002] Chromatography is a common instrument for analyzing complex compounds. Retention time refers to the time from the beginning of injection to the maximum value of the chromatographic signals corresponding to various components, and is an important indicator to distinguish various components. In the multi-batch experimental design, the retention time of the same component in different samples will deviate due to the change of the instrument state. The method of correcting for this bias is called retention time alignment. Quantitative proteomics experiments are mainly based on liquid chromatography-mass spectrometry (LC-MS). The mass spectrometer is a multi-stage tandem mass spectrometer. The traditional ...

AI Technical Summary

Problems solved by technology

Some work has used deep learning for the alignment of gas chromatography retention time, but the characteristics of its deep learning network use secondary spectra, which limits its application range

Images