Fig. 9
From: ScLRTC: imputation for single-cell RNA-seq data via low-rank tensor completion
Violin chart of data expression distribution and accuracy measurements of DE genes by scLRTC and other methods. (A) Violin chart of data expression distribution after imputation when the dropout rate is 40% (data is transformed by log10(X + 1)). The more similar the shape of the violin is to FULL, the more effective the imputation effect. (B) ROC curves and AUC scores of DE genes with different imputation methods. AUC combines the recall rate and precision rate, and the value closer to 1 indicates a better imputation method. Here, the recall rate is defined as the number of true positives divided by the total number of samples that actually belong to the positive class, and the precision rate is the number of true positives divided by the total number of samples labelled as belonging to the positive class