Fig. 7: Prediction of association between gene expression and peak accessibility of transcription factors. | Nature Communications

Fig. 7: Prediction of association between gene expression and peak accessibility of transcription factors.

From: multiDGD: A versatile deep generative model for multi-omics data

Fig. 7

A Violin plots of fractions of closing peaks (∆X < 0) for in silico silencing of transcription factors (TFs) annotated as activators (red, 32895 cells, 34 TFs) or repressors (blue, 7226 cells, 7 TFs). The black box and white point denote the interquartile range and median of the distribution, respectively. The p-value for a two-sided 2-sample T-test comparing the mean of the distributions is shown on top. B Mean ∆X in response to TF silencing for 34 predicted activators and 7 annotated repressors. We show the mean effect for all affected cells over a set of 10,000 peaks containing TF binding motifs (red points) or over a set of 10,000 peaks sampled at random amongst the peaks containing at least one TF binding motif (gray points) with the standard error of the mean. Asterisks denote TF perturbations for which the difference in mean effect between TF binding peaks and random peaks is significant (two-sided 2-sample T-test p-value < 0.01). All values are provided in the Source Data.

Back to article page