Benchmarking the Autoencoder Design for Imputing Single-Cell RNA Sequencing Data

This repository contains the real and synthetic datasets used in the paper "Benchmarking the Autoencoder Design for Imputing Single-Cell RNA Sequencing Data". The zip file includes three folders:

1. overall imputation accuracy: the 12 real scRNA-seq datasets used in the evaluation of overall imputation accuracy.

2. cell clustering: the 20 real scRNA-seq datasets with cell type labels used in the evaluation of cell clustering.

3. DE gene: the 20 scRNA-seq syntehtic datasets with ground-truth DE genes used in the evaluation of DE gene analysis. These datasets are simulated by simulator scDesign and 20 real datasets.