Single-cell transcriptome clustering method, system and device based on deep autoencoder

CN122333008APending Publication Date: 2026-07-03ANHUI UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
ANHUI UNIV
Filing Date
2026-06-05
Publication Date
2026-07-03

AI Technical Summary

Technical Problem

Existing single-cell RNA sequencing data clustering methods suffer from high-dimensional sparsity, high noise levels, batch effect correction sacrificing biological variation, low recall for rare cell type identification, high computational complexity, and a lack of effective methods to distinguish between technical zero values ​​and biological zero expression.

Method used

A deep autoencoder-based clustering method is adopted. Through data preprocessing, deep autoencoder network construction, design of multi-task loss function, model training and feature extraction, dimensionality reduction visualization and spectral clustering analysis, combined with ZINB reconstruction loss and contrastive loss, UMAP dimensionality reduction and spectral clustering algorithm are used for cell clustering.

Benefits of technology

It improves the clustering accuracy of single-cell data, enhances the discriminative power of cell types, effectively handles zero expansion characteristics and high noise, identifies complex cell types, and has good scalability.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122333008A_ABST
    Figure CN122333008A_ABST
Patent Text Reader

Abstract

This invention discloses a method, system, and device for single-cell transcriptome clustering based on deep autoencoders. The method includes: data preprocessing and feature engineering, construction of a deep autoencoder network, design of a multi-task loss function, model training and feature extraction, dimensionality reduction visualization, and spectral clustering analysis. The system includes: a data preprocessing module, a deep autoencoder module, a loss function calculation module, a model training module, a dimensionality reduction visualization module, a clustering analysis module, and an evaluation module. This invention significantly improves the clustering accuracy of single-cell data by introducing a contrastive learning mechanism and optimizing the clustering strategy.
Need to check novelty before this filing date? Find Prior Art