ANCOVA predicting r values for miRNA – host gene co-expression values (r) as a function of whether they are conserved (dashed regression line, open squares) or non-conserved (solid regression line, closed circles), controlling for abundance of the miRNA (logged). In all cases a prior test was run to check that the interaction term between the predictor variables was not significant as such an interaction renders an ANCOVA invalid. In all cases the interaction was non-significant. The P value indicates the significance of the hypothesis that the conserved genes have a higher r than non-conserved ones. In all instances it is significant. Data is the same as employed for Table 3. The order of sample sets for the Supplementary Figures is leukemia set (Fig. S1), myeloma set (Fig. S2), lymphoblastoid cell set (Fig. S3), breast cancer set (Fig. S4), and prostate set (Fig. S5).
http://postimg.org/image/wj3o7vqsb/