Considering structural and functional data, we looked at perturbation values per position, and considered the percentage of positive scores and negative scores, keeping the maximum of the two.

This presents a measure of consensus between the changes at each position, the higher values represent that most mutations result in the same effect (whether positive or negative values), independent of the mutant amino acid or the amino acid being replaced. We present averages and standard deviations over positions and proteins for the nodes, edges, and weight measures (structural data) and for the functional data.