Figure 4.

Total number of predicted proteins that show significant similarity (> 20% identity to one partner and > 10% less identity to the other) over their entire length (grey bar) or within domains of at least 50 amino acids (black bar).