Representing fusion proteins

[cthoyt]

Every once and a while, I run into something like

LPL–GPIHBP1 fusion protein showed high enzymatic activity in in vitro assays using surrogate substrates as well as the natural LPL substrates VLDL and CM.

that gets picked up as a binding (i.e., a complex statement), but being a fusion protein is a totally different kind of phenomena. There are a few things that could help reduce curation burden on these:

1. Add some simple rules for filtering these out from complexes, since fusion protein usually is pretty explicit
2. Add new rules to reach?
3. Consider creating a new statement type to represent this explicitly