Move AD/PD networks to supplementary materials

[bschilder]

There's already quite a few main figs and added the AD/PD figures might be a bit much. @NathanSkene suggested they dont go thematically with the rest of the paper, which is true in some senses since these examples were handpicked by me because of our lab's interests and also the broader interest in these diseases.

Changes to the severity scoring (eg #68) will affect which phenotypes are the top prioritised targets. We should pick phenotypes that were highly ranked in the prioritisation pipeline, but the phenos to show in the network plots also need to consider which ones are most visually interpretable (thus why I picked Tay-Sachs and lethal skeletal dysplasia which have relatively simple networks).

[bschilder]

current version: