
Patient subnetwork analysis highlights core KMT2A-FP GRN circuitry. (A) Subnetworks reflecting expressed genes in patient samples. (B) Percentage of samples in each RNA-seq data set that express each gene plotted against degree centrality in the KMT2A-AFF1 GRN. (C) UMAP dimensionality reduction of node activity across subnetworks, colored according to k-means clusters. (D) Heatmap of node activity across subnetworks, separated into k-means clusters as in C. Rows represent genes; columns represent samples. Red and gray indicates gene expression above and below the mean, respectively. (E) Violin and box plots showing distribution of log2 degree centrality of GRN nodes, stratified by k-means cluster. Box plot midline indicates median, and upper and lower hinges indicate 1st and 3rd quartiles, respectively. Whiskers extend to furthest values within 1.5 times the interquartile range. (F) Degree centrality of nodes plotted against rank order of centrality. Annotated nodes in Clusters 1, 2, and 4 represent the core TFs of the GRN model.











