Lingering issues and concerns in inclusion criteria

Resolution of some lingering conflicting annotations. RATT and ab initio annotations that overlap in-frame but are differently named should already be treated as conflicting and only one can pass. However,

• ab initio annotations that are unnamed and overlap in-frame with a pseudo RATT annotation are also passing, but they shouldn't:

PAK     2554222 2554539 1               ORF0276 Prokka  FALSE
PAK     2554329 2554539 1               PA2668  RATT    TRUE

PAK     2929227 2930301 -1              vgrG3   RATT    TRUE
PAK     2929227 2931230 -1              ORF0029 Prokka  FALSE

PAK     4882806 4883952 -1              hxcY    RATT    TRUE
PAK     4882806 4884017 -1              ORF0089 Prokka  FALSE

• When allowing complete postprocessing of ab initio genes, leaving resolution of genes with identical coordinates and different names to check_inclusion_criteria, they are not treated as mutually exclusive annotations when one is pseudo. That was the original design, but it might need to be revisited for such stark conflicts:

AZ20    62253   63716   1               ECOLIN_25415    RATT    TRUE
AZ20    62334   63716   1               ECOLIN_20185    Prokka  FALSE

AZ20    3673193 3673651 -1              ECOLIN_02595    RATT    TRUE
AZ20    3673193 3673651 -1              ECOLIN_02600    Prokka  FALSE

...

closes #25 (closed) #51 (closed) #57 (closed)