scASGC: An adaptive simplified graph convolution model for clustering single-cell RNA-seq data

Shudong Wang, Yu Zhang, Yulin Zhang, Wenhao Wu, Lan Ye, YunYin Li, Jionglong Su, Shanchen Pang

Abstract: Highlights•Remove noise in the similarity matrix by Network Enhancement to build more reliable cell graphs.•Aggregate neighborhood information using a simplified graph convolution model to make full use of the relationships between cells, as well as greatly improve the efficiency of the model.•Selecting the appropriate number of convolution layers adaptively for different data allows flexible response to a variety of data.•Outperforms existing methods in terms of clustering accuracy and efficiency on 12 publicly available scRNA-seq datasets.