Figure 1. Protein family counts for the ten possible informative profiles.

The table was adapted from Lake's Table 1 [2] to include the Pfam counts that result if different representative classes are chosen for the DM group. Number of Pfam per group is in parentheses the same number as in Lake's paper was found for all other groups. The circle illustrates Lake's hypothesis that the double membrane bacteria resulted from a fusion between Clostridia and Actinobacteria. The patterns compatible with this hypothesis are boxed (pattern 5,7,8,9 and 10).