Fig. 1From: Comparison of Scanpy-based algorithms to remove the batch effect from single-cell RNA-seq dataBatch-corrected results for lung data from MCA and TM. a, The t-SNE plots present the degree of the batch effect from the MCA lung data (consisting of 3 experimental batches) before correction (baseline) and after correction with 4 methods (Regress_Out, ComBat, Scanorama and MNN_Correct). b, c, ASW_batch (boxplot) and the kBET rejection rate (line chart) evaluate the batch-correction effect in the MCA lung data. d, The t-SNE plots present the degree of the batch effect from the TM lung data (consisting of 4 batches) before correction (baseline) and after correction using the 4 methods (Regress_Out, ComBat, Scanorama and MNN_Correct). e, f, ASW_batch (boxplot) and the kBET rejection rate (line chart) evaluate the batch-correction effect in the TM lung data. *p < 0.05, **p < 0.01, ***p < 0.001; the Wilcoxon signed-rank test with Benjamini and Hochberg correction was performed between each of the four postcorrection groups and the baseline groupBack to article page