Modify the ancestral inference to be able to infer on per-chromosome basis

[janaobsteter]

Currently, the ancestral pipeline runs on the whole vcf at once. I think we should add a modification to be able to run by chromosome if needed for large datasets.

[janaobsteter]

The solution is in #61