Fig 2: Identification and characterization of different EMT states in HCC cell lines(A) Scatter diagram of epithelial (E) and mesenchymal (M) states for five HCC cell lines at the single-cell level (top panel) and pseudo-bulk level (bottom panel) based on single-cell RNA sequencing (scRNA-seq) data. The top left of the arrow indicates M state, and the bottom right indicates E state. In the pseudo-bulk level analysis, the average scores of E and M in each cell line were plotted on the scatter diagram.(B) Density distribution of E (pink) and M (light blue) scores in five HCC cell lines.(C) Density map of normalized GA based on scATAC-seq data. CDH1, a classic epithelial gene. KRT14, an early EMT gene. SNAI2 and TWIST1, classical EMT transcription factors.(D) EMT scatter diagram showing the relevance of adhesion and migration related surface markers (CD24, CD44, EPCAM, and CD54) with E or M state using normalized protein expression. p and r values of Pearson correlation coefficients analysis were compiled in Table S3.(E) Violin plot depicting EMT state in each cell line using the normalized expression level of CD24, EPCAM, CD44, and CD54 in triple-omics. Wilcoxon rank-sum test was used to test the statistically significant differences. p values were compiled in Table S4.(F) EMT scatter diagram showing the normalized protein expression level of CD24, EPCAM, CD44, and CD54 in Huh7. Cells with expression greater than the upper quartile were defined as “High”, less than the lower quartile were defined as “Low”, and others were defined as “Medium”. p and r values of Pearson correlation coefficients analysis were compiled in Table S3. See also Figures S1 and S2 and Tables S2–S4.