Draft: Removing genomes from an existing pangenome.

Debugging still in progress.

Tested on chloroplasts datasets: add C_sativus then Z_mays; then remove Z_mays.

Currently export comparison:

• anchors: correct
• nodes: correct except pangenome node: a few properties are not correctly updated; num kmers will be an issue, because kmers of Z_mays get node index -1L, instead of being removed from the index database (maybe keep track of kmers that would have been removed and reduce pangenome property by that).
• relationships: incorrect: seems to be a repetition problem where all positions on the remaining genome are put in there publicate (maybe only for some edges).

TODO:

• test on more chloroplasts, also testing removing genome other than last one.
• test on hiv datasets (Sandra).
• discuss future steps for handling removed genomes and the 'empty space' they leave in some listed (array) properties.

requested review from @worku005

assigned to @lankveld_t

I think the current problem with the relationships may be due to running localize_nodes() near the end of the process (mostly as a quick way to get the anchors right). I think this may only add info to the edges, instead of removing it first.