ScRNA-seq data dimension reduction method based on deep adversarial variational auto-encoder

Abstract

The invention relates to data mining in bioinformatics, in particular to mining of single-cell RNA sequencing data and provides an ScRNA-seq data dimension reduction method based on a deep adversarial variational auto-encoder. specifically, dimension compression and clustering are carried out on single-cell RNA sequencing data through a deep learning method, so that the purpose of effectively identifying cell populations is achieved. The method disclosed by the invention comprises the following steps of: collecting scRNA-seq data and preprocessing the scRNA-seq data; constructing the deep adversarial variational auto-encoder model; carrying out dimension reduction on the preprocessed data by using the constructed model; combining the deep adversarial variational auto-encoder with a Bhatpacaryya distance; and performing clustering analysis on a result obtained dimension reduction. The model restrains a data structure, and dimension reduction is carried out through the deep adversarial variational auto-encoder module. Based on an evaluation index, namely standardized mutual information, experiments carried out on three real scRNA-seq data sets show that the method has good performance.

Description

technical field [0001] The invention relates to data mining in bioinformatics, in particular to the mining of single-cell RNA sequencing data. Specifically, it involves dimensionality compression and clustering of single-cell RNA sequencing data to achieve the purpose of effectively identifying cell populations. Background technique [0002] Single-cell RNA sequencing (scRNA-seq) is not a surface understanding of only sequencing a single cell. We can understand that this sequencing technology can sequence the genome or transcriptome of a single cell, that is, sequencing at the single-cell level. Traditional sequencing methods are generally performed at the level of multiple cells, thus losing the information of heterogeneity. Rather, the information revealed by conventional sequencing methods is averaged at the multicellular level. [0003] With the development of scRNA-seq technology, a large amount of scRNA-seq data has been generated, which provides strong data support ...

AI Technical Summary

Problems solved by technology

The method of the present invention can effectively solve the problems of important information loss and insufficient feature extraction existing in existing dimensionality reduction methods, and obtain better clustering accuracy

Images