Fig. 5: High-throughput gene expression-based combinatorial drug screen. | Nature Communications

Fig. 5: High-throughput gene expression-based combinatorial drug screen.

From: Combi-seq for multiplexed transcriptome-based profiling of drug combinations using deterministic barcoding in single-cell droplets

Fig. 5

a TSNE plots of normalized gene expression data. YM155 (left panel), Blebbistatin (middle panel), and YM155-Blebbistatin combination treated samples are labeled as representative examples. b Silhouette scores of sample clustering based on autosampler/Braille valves drugs and combinations. Silhouettes scores are compared to random distributions (color code) created by permuting sample labels. (c) ROC analysis of drug signature similarity to LINCS-L1000 data. For each drug, a consensus signature was calculated and similarity (Spearman’s rank correlation) to corresponding LINCS-L1000 signatures was calculated. The similarity values were used as predicted values for ROC analysis, while true positives were the matched drug pairs between the high-throughput screen and LINCS-L1000. Source data are provided as a Source Data file.

Back to article page