Journal: bioRxiv
Article Title: BARTsc identifies key transcriptional regulators from single-cell omics data
doi: 10.64898/2026.02.24.707729
Figure Lengend Snippet: ( a ) UMAP visualization of the scMultiome dataset for human PDAC, where each point represents a single cell colored by cell type. ( b ) Signature analysis results for selected TRs in human PDAC. The heat map is colored by BARTsc score. Asterisks indicate significance level, *: p < 0.05, **: p < 0.01, ***: p < 0.001. ( c , d ) Performance comparison between BARTsc and other methods in identifying key regulators for each cell type from PDAC scMultiome dataset. Bar plots show the F1 scores for key regulator prediction for different methods across different cell types ( c ). Scatter plot shows the sensitivity and specificity of each method in identifying cell type key regulators ( d ). ( e - h ) UMAP visualizations of the ductal cells from the PDAC scMultiome dataset, where each point represents a single ductal cell colored by sub-clusters ( e ), or the normalized expression level of gene MKI67 ( f ), TOP2A ( g ) and CDK1 ( h ). ( i ) Scatter plot showing TRs ranked by BARTsc score. The top five predicted key regulators are highlighted and labeled. ( j ) Chronos scores quantifying the effects of CRISPR-mediated knockout of the top five predicted key regulators in PDAC cell lines. Each point represents a PDAC cell line, with lower Chronos scores indicating greater loss of cell fitness upon gene knockout. P-values are calculated by one-sided Wilcoxon rank-sum tests. ( k ) RT-qPCR results showing the relative mRNA levels of NELFA and selected genes upon NELFA knock-down by siRNA in PANC-1 cells. Bars represent the mean mRNA levels relative to those under control siRNA transfection groups after normalization to GAPDH mRNA levels, and error bars indicate the standard deviation. Asterisks indicate significance levels between siNELFA and siControl determined by one-sided t -test, *: p < 0.05, **: p < 0.01, ***: p < 0.001. ( l ) Growth of PANC-1 cells three days after NELFA knockdown (siNELFA) compared with the control cells (siControl). Error bars indicate the standard deviation. P-value is calculated by one-sided t -test.
Article Snippet: Human pancreatic carcinoma cell line PANC-1 was originally purchased from the American Type Culture Collection (ATCC) (Cat# CRL-1469).
Techniques: Single Cell, Comparison, Expressing, Labeling, CRISPR, Knock-Out, Gene Knockout, Quantitative RT-PCR, Knockdown, Control, Transfection, Standard Deviation