Advertisement

论文阅读：Deep structural clustering for single-cell RNA-seq data jointly through AE and GNN

阅读量：

Yanglan Gan, Xingyu Huang, Guobing Zou, Shuigeng Zhou, Jihong Guan, Deep structural clustering for single-cell RNA-seq data jointly through autoencoder and graph neural network, Briefings in Bioinformatics , Volume 23, Issue 2, March 2022, bbac018, Deep structural clustering for single-cell RNA-seq data jointly through autoencoder and graph neural network | Briefings in Bioinformatics | Oxford Academic

论文地址：Deep structural clustering for single-cell RNA-seq data jointly through autoencoder and graph neural network | Briefings in Bioinformatics | Oxford Academic (oup.com)

代码地址：DHUDBlab/scDSC (github.com)

摘要

单细胞RNA测序（scRNA-seq）可以研究细胞异质性和多样性的复杂机制。无监督聚类对于分析scRNA-seq数据至关重要，因为它可以用于识别潜在的细胞类型。然而，由于噪声影响、高维度和普遍的缺失事件，scRNA-seq数据的聚类分析仍然是一个计算挑战。本文提出了一种新的用于scRNA-seq数据的深度结构聚类方法，命名为scDSC，它将结构信息整合到单细胞的深度聚类中。所提出的scDSC包括基于零膨胀负二项式（ZINB）模型的自编码器，图神经网络（GNN）模块和相互监督模块。为了从稀疏和零膨胀的scRNA-seq数据中学习数据表示，将ZINB模型添加到基本自编码器（AE）中。引入GNN模块以捕获细胞之间的结构信息。通过将基于ZINB的自编码器与GNN模块相结合，模型将自动编码器学习到的数据表示转移到相应的GNN层。此外，采用相互监督策略来统一这两种不同的深度神经架构，并引导聚类任务。

模型图

看过SDCN模型的同学，可以发现这一篇的思路就是将SDCN模型用于单细胞RNA测序任务。为了适应单细胞RNA数据的复杂分布，将常规的AE改为基于ZINB的AE。

论文的写法也与SDCN论文类似。如果想把深度聚类模型融入该领域中，可以借鉴一下该文的切入角度和实验角度。

全部评论 (0)

还没有任何评论哟~

相关文章推荐

论文阅读：Deep structural clustering for single-cell RNA-seq data jointly through AE and GNN

YanglanGan,XingyuHuang,GuobingZou,ShuigengZhou,JihongGuan,DeepstructuralclusteringforsinglecellRNAse...

论文阅读：scDFC: A deep fusion clustering method for single-cell RNA-seq data

DayuHu,KeLiang,SihangZhou,WenxuanTu,MengLiu,XinwangLiu,scDFC:Adeepfusionclusteringmethodforsinglecel...

论文阅读：scGAC: a graph attentional architecture for clustering single-cell RNA-seq data

YiCheng,XiuliMa,scGAC:agraphattentionalarchitectureforclusteringsinglecellRNAseqdata,Bioinformatics,...

论文阅读：scCAN: Clustering With Adaptive Neighbor-Based Imputation Method for Single-Cell RNA-Seq Data

S.Dong,Y.Liu,Y.Gong,X.DongandX.Zeng,scCAN:ClusteringWithAdaptiveNeighborBasedImputationMethodforSing...

【论文阅读】 Single-cell RNA-seq denoising using a deep count autoencoder.

简介 这是一篇2019年1月23日发表在naturecommunications杂志上的文章。文章主要介绍了一种使用DCA对单细胞count数据进行降噪，其中自编码器的loss设计是本篇文章的亮点。

【论文阅读】Integrating single-cell multi-omics data through self-supervised clustering

论文地址：IntegratingsinglecellmultiomicsdatathroughselfsupervisedclusteringScienceDirect 代码地址：https://gi...

【论文阅读】Co-clustering of single-cell RNA-seq data based on weighted non-negative matrix tri-factoriza

摘要 单细胞RNA测序（scRNAseq）能够以单细胞分辨率精确识别组织中包含的细胞类型。然而，利用scRNAseq数据识别不平衡细胞类型的研究仍然具有挑战性。

论文阅读：ZINB-Based Graph Embedding Autoencoder for Single-Cell RNA-Seq Interpretations

Yu,Z.,Lu,Y.,Wang,Y.,Tang,F.,Wong,K.C.,&Li,X.2022.ZINBBasedGraphEmbeddingAutoencoderforSingleCellRNAS...

【论文阅读】scCAT: Single-cell Combined graph Attentional clustering for scRNA-seq analysis

论文地址：scCAT:SinglecellCombinedgraphAttentionalclusteringforscRNAseqanalysisScienceDirect 摘要 单细胞RNA测序（...

【论文阅读】Attention-based deep clustering method for scRNA-seq cell type identification

论文地址：AttentionbaseddeepclusteringmethodforscRNAseqcelltypeidentificationPLOSComputationalBiology 代码地...