Fig. 3From: Learning a hierarchical representation of the yeast transcriptomic machinery using an autoencoder modelMapping between transcription factors (TFs) and hidden variables in the first hidden layer. Results for sparse autoencoder (a) and NMF (b) are shown. The transcription factors (TFs) are arranged along x-axis, and the hidden variables are arranged along y-axis. Each point in the figure represents the value of –log(p-value) of the enrichment score between genes regulated by a hidden node and a TF. The pseudo-color bar shows the scale of the –log(p-value)Back to article page