Sorting of sample IDs should take pathways into account

[selkamand]

We should sort patients after we've attached pathway to the data.frame - We want to sort so that mutations up the top of the oncoplot are left-most - then have our code before that simply decide which genes/pathways we want up the top of the oncoplot. by default these can be based on how heavily mutated the pathway/sample is - but if we change how the factors are releveled to put new genes/pathways up the top - sample-level sorting should automatically sort samples accordingly