CL-Impute: A contrastive learning-based imputation for dropout single-cell RNA-seq data

Yuchen Shi, Jian Wan, Xin Zhang, Yuyu Yin

Abstract: Highlights•An approach to recover missing genes due to dropout events on scRNA-seq data.•CL-Impute can learn cell features from self- and global-perspective, respectively.•CL-Impute outperforms other methods in downstream analysis of scRNA-seq data.•Demonstrate excellent performance in the context of high dropout rates.•CL-Impute can explore the nondropout cell features by receiving dropout inputs.