Large-scale single cell typing method and system and storage medium

Abstract

The invention provides a large-scale single cell typing method and system and a storage medium, and the method comprises the steps: obtaining a gene sample, screening the gene sample to obtain a high-variation gene, and constructing a training data set according to the high-variation gene; training a deep learning model according to the training data set; and performing cell typing through the trained deep learning model, wherein according to the training data set, training of the deep learning model comprises the following steps: performing sample grouping according to an Euclidean distance and the training data set to obtain a first grouping sample; performing linear scaling on the first grouping sample to obtain a second grouping sample; and carrying out dimension reduction processing on the second grouping sample through an auto-encoder in combination with prior biological knowledge. The method reduces the influence of technical variability on the real biological difference, can provide powerful evidence for comprehensively revealing the characteristics and functions of various cells, and can be widely applied to the technical field of bioinformatics.

Description

technical field [0001] The invention relates to the technical field of bioinformatics, in particular to a large-scale single cell typing method, system and storage medium. Background technique [0002] The most basic unit of organisms is the cell, and human beings consist of approximately 3.72×10 13 composed of cells. These cells work in harmony to maintain the normal operation of the body. However, as long as a single cell becomes cancerous and continues to proliferate, it may destroy the entire living body. Existing research at the transcriptome level is to sequence large tissues containing hundreds or thousands, and it is very difficult to find a small number of mutant cells in them, so biologists need a new method to study cells. Single cell RNA sequencing (Single cell RNA-Seq, scRNA-seq) is a method to analyze its transcripts at the level of a single cell. The first article on single-cell RNA sequencing was published in 2009. Since its inception, single-cell RNA seq...

AI Technical Summary

Problems solved by technology

For example, single-cell RNA sequencing uses higher amplification multiples and deeper sequencing depth, so some errors may be amplified, such as GC content may affect the accuracy of sequencing

In addition, since single-cell RNA sequencing only measures the transcription level of a single cell, compared with mass tissue sequencing, its cell-specific expression signals are likely to be concealed by differences in experimental conditions

Images