PTC sequence and endothelial contexts. As for other figures, ACVRL1+/− is highlighted in pink/purple; ENG+/− in blue, and SMAD4+/− in orange. (A) Unsupervised Euclidean distance analyses for the 16 BOEC RNA sequencing data sets: note efficiency of PTC loss was not associated with the differences identified by the initial global expression profiles. (B) BOEC expression of the 25 lowest Gini coefficient genes61,62 (GC < 0.15) with GC < 0.13 genes highlighted in green (details in supplemental Table 2; supplemental Figure 8). (C) Unsupervised analyses of the 24 BOEC-expressed GINI genes ranked in 16 BOEC cultures. (i) Multivariate analysis of genes (observations, gray dots) across the BOECs (variables, green arrows). Note demarcation of the 4 ENG+/PTC BOEC cultures. (ii) Principal factor analysis on 16 BOEC data sets using ranks of the 24 GINI genes. The 15 retained factors and 120 evaluated parameters again distinguished the ENG+/PTC BOEC cultures (blue) from ACVRL1+/PTC (pink); SMAD4+/PTC (orange); and control BOECs (gray). (D) Respective ranks of the 5 lowest GC genes (highlighted green in panel B) across BOEC cultures ranked by % PTC persistence and genotype (C: control, E: ENG+/PTC, S: SMAD4+/PTC, A: ACVRL1+/PTC). Note switch between 14% and 11% persistence for NXF1 (blue) and RBM45 (brown). Two-way analysis of variance (ANOVA) data are presented in supplemental Table 7.
