Pairing subgrapHs Using NetworK Environment Equivalence

COG function matching uses both sequence similarity and function information to assign node matches. Known functions for each protein are taken from the COG database.

For each COG function present in both networks, a node match is assigned to the most sequence-similiar protein pair with that function. This ignores matches between sequence-similar proteins of unknown function and multiple protein pairs with the same function. However, it does ensure that the resulting matches are consistent in terms of function. This is useful to gauge the impact of inaccurate matches resulting in other methods.