Skip to main content
Fig. 3 | BMC Bioinformatics

Fig. 3

From: Benchmarking imputation methods for network inference using a novel method of synthetic scRNA-seq data generation

Fig. 3

Synthetic scRNA-seq data generated for 5 gene network example. The network was simulated using 500 cells over a 500 hour time period with parameters sampled as described in Methods sect. "Parameters for mammalian cells" A Plot of the volume time series of a single representative cell. Early divisions are due to replacement in order to keep number of tracked cells constant. B Plot of corresponding mRNA time series for the 5 genes modelled. C Plot of corresponding protein time series for the 5 genes modelled. D Heatmap of mRNA pearson correlations taken from final time point. E Heatmap of mRNA pearson correlations scaled by cell volume taken from final time point. F Heatmap of mRNA pearson correlations scaled by cell volume and subsequently downsampled using Binomial downsampling with 20% capture efficiency. G Graph of ground truth network where a blue arrow represents a link with an activating reaction and an orange arrow represents a link an inhibiting reaction. H Graph of inferred reaction network obtained from PIDC algorithm using mRNA data scaled by cell volume at final time point as input. Predicted links are represented by solid black lines. I Graph of inferred reaction network obtained from PIDC algorithm using mRNA data scaled by cell volume at final time point and downsampled (using Binomial downsampling with 20% capture efficiency) as input. Predicted links are represented by solid black lines

Back to article page