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

Single-cell RNA sequencing (scRNA-seq) data can be a good indicator of cell-to-cell heterogeneity and can help to study cell growth by identifying cell types. With advances in sequencing technology, single-cell RNA data are rapidly accumulating, and the large and complex data require new algorithms to extract the valid information contained in the integrated data. Deep learning has proven to be effective and innovative in processing complex data. Therefore, incorporating deep learning algorithms to study single-cell RNA data is a leading edge and hot topic of current research. Deep clustering is a combination of deep learning and conventional clustering, which trains embedding and clustering together to obtain the optimal embedding subspace for clustering, which is more effective compared to conventional clustering methods. 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.