Skip to main content
Fig. 2 | Genome Biology

Fig. 2

From: Integration of single-cell multi-omics data by regression analysis on unpaired observations

Fig. 2

Performance of UnpairReg in silico mixture of cells. A PCC of UnpairReg predicted coefficient beta and real beta under different dropout rates. B The mean cell-level PCCs between predicted gene expression level and real gene expression. Cell-level PCC is PCC calculated for each cell across genes. Color in B to E indicates the different types of predicted gene expression data. Orange represents for UnpairReg predicted gene expression; blue represents observed gene expression after drop-out; grey represents random data. C The average of gene-level PCCs between predicted gene expression level and ground truth. Gene level PCC is PCC calculated for each gene across cells. D Cell-level PCC for each cell under a dropout rate of 0.87. The value labeled is the -log10 p-value of the one-sample t-test. E Gene level PCC under a dropout rate of 0.87. The value labeled is the same as D. Dropout rate is defined by the percentage of 0 in the single-cell data

Back to article page