Fig. 4: Assessing the robustness, sensitivity, and accuracy of Combi-Seq.

a Heatmap of correlations between pathway activities of bulk and droplet Combi-Seq data: PROGENy pathway activities were determined for each sample and correlations for matched samples using either data obtained in microtiter plates (rows) or droplets (columns) were calculated (color code blue to red) and used to perform hierarchical clustering. Drugs of combinations are color-coded (light green: Autosampler drug, blue: Braille valves drug, cyan: Same drug from autosampler and Braille valves). b Accuracy of spike-in detection as a correlation between ERCC input concentration and measured transcripts per million (TPM) for 250 cells per sample, shaded area shows 95% confidence interval of the regression estimate. (c) Summary of correlation coefficients for 125 (R = 0.63 and R² = 0.40), 250 (R = 0.65 and R² = 0.42) and 500 cells (R = 0.7 and R² = 0.49) per sample, n = 5 biological independent experiments, data are presented as mean values ± 95% confidence interval. d Sensitivity as the detection probability for spike-in molecules for 250 cells per sample. A spike-in was considered as detected when the probability reached 50%. e Summary of sensitivities for different amounts (125, 250, and 500) of cells per sample, n = 5 biological independent experiments, data were presented as mean values ± 95% confidence interval. Source data are provided as a Source Data file.