Fig. 6

Single-cell gene regulatory network inference reveals candidate regulators of XCR during iPSC reprogramming. A tSNE visualizations of single-cell clustering based on regulon activity. Each dot represents a cell. Top tSNE: colors indicate isolation time point (pink = day 0, blue = day 8, green = day 9, yellow = day 10, red = day 12, and brown = iPSCs) with cell states marked with dashed lines (red = somatic, yellow = intermediate and blue = pluripotent). Bottom tSNE: colors indicate graph-based clustering classification (red = C0, yellow = C1, green = C2, light blue = C3, dark blue = C4, and pink = C5). Clustering based on regulon activity was performed with SCENIC on Smart-seq2 dataset. The activity of 311 regulons in total was quantified. B Heatmap of regulon activity ordered by cell state and pseudotime (x-axis). Dashed lines indicate cell states (red = somatic, yellow = intermediate, and blue = pluripotent). Selected regulons are indicated in Y-axis and regulon activity of selected regulons is shown in tSNEs (right) with the corresponding motif. C Heatmap with regulon activity of regulons with the highest specificity for each cluster (C0–C5). D Plot with the number of X-linked targets per regulon. Top 20 regulons with the most X-linked targets are shown. E Relationship between regulon activity and XCR. Four regulons are displayed with the highest coefficients of logistic regression model of X-linked genes allelic ratio on regulon activity. Fitted line on the plot was generated using loess function. The grey areas represent the 95% confidence interval