ScInfoVAE: interpretable dimensional reduction of single cell transcription data with variational autoencoders and extended mutual information regularization

BioData Mining

Research

Faning Long¹  Weiquan Pan²  Jian Pan² 

[1] School of Computer Science and Engineering, Yulin Normal University, Yulin, China;School of Mathematics and Statistics, Yulin Normal University, Yulin, China;

关键词: Unsupervised clustering;    Deep neural network;    Variational autoencoder;    Mutual information;    scRNA-seq data;   

DOI  :  10.1186/s13040-023-00333-1

 received in 2022-03-23, accepted in 2023-06-05,  发布年份 2023

来源: Springer

【 摘 要 】

Single-cell RNA-sequencing (scRNA-seq) data can serve as a good indicator of cell-to-cell heterogeneity and can aid in the study of cell growth by identifying cell types. Recently, advances in Variational Autoencoder (VAE) have demonstrated their ability to learn robust feature representations for scRNA-seq. However, it has been observed that VAEs tend to ignore the latent variables when combined with a decoding distribution that is too flexible. In this paper, we introduce ScInfoVAE, a dimensional reduction method based on the mutual information variational autoencoder (InfoVAE), which can more effectively identify various cell types in scRNA-seq data of complex tissues. A joint InfoVAE deep model and zero-inflated negative binomial distributed model design based on ScInfoVAE reconstructs the objective function to noise scRNA-seq data and learn an efficient low-dimensional representation of it. We use ScInfoVAE to analyze the clustering performance of 15 real scRNA-seq datasets and demonstrate that our method provides high clustering performance. In addition, we use simulated data to investigate the interpretability of feature extraction, and visualization results show that the low-dimensional representation learned by ScInfoVAE retains local and global neighborhood structure data well. In addition, our model can significantly improve the quality of the variational posterior.

【 授权许可】

CC BY   
© The Author(s) 2023

【 图 表 】

40517_2023_259_Article_IEq86.gif

40517_2023_259_Article_IEq52.gif

Fig. 4

40517_2023_259_Article_IEq51.gif

40517_2023_259_Article_IEq90.gif

40517_2023_259_Article_IEq89.gif

40517_2023_259_Article_IEq48.gif

40517_2023_259_Article_IEq85.gif

40517_2023_259_Article_IEq84.gif

40517_2023_259_Article_IEq46.gif

40517_2023_259_Article_IEq45.gif

Fig. 2

Fig. 1

40517_2023_259_Article_IEq42.gif

40517_2023_259_Article_IEq41.gif

Fig. 1

40517_2023_259_Article_IEq35.gif

Fig. 2

40517_2023_259_Article_IEq32.gif

40517_2023_259_Article_IEq31.gif

40517_2023_259_Article_IEq30.gif

Fig. 1

40517_2023_259_Article_IEq28.gif

40517_2023_259_Article_IEq27.gif

Fig. 2

Fig. 1

42004_2023_919_Article_IEq176.gif

42004_2023_919_Article_IEq175.gif

Fig. 2

Fig. 8

Fig. 5

42004_2023_919_Article_IEq171.gif

Fig. 4

Fig. 1

42004_2023_919_Article_IEq168.gif

13011_2023_540_Article_IEq40.gif

40517_2023_259_Article_IEq25.gif

Fig. 1

40517_2023_259_Article_IEq19.gif

40517_2023_259_Article_IEq18.gif

40517_2023_259_Article_IEq17.gif

40517_2023_259_Article_IEq16.gif

40517_2023_259_Article_IEq15.gif

Fig. 3

40517_2023_259_Article_IEq7.gif

13011_2023_540_Article_IEq19.gif

Fig. 1

13011_2023_540_Article_IEq1.gif

41116_2023_37_Article_IEq286.gif