See if applying parsimony to the sites changes the percentage of single mut sites

[hyanwong]

We say something like this:

Many additional sites contain multiple mutations, albeit with less extreme counts than in these outliers. In total, out of 1,365,881 non-singleton variant sites on Chromosome 20, 825,578 (60%) host more than one mutation. However, the more samples we have, the higher the chance of a sequencing error in at least one of the samples, likely to appear in the tree sequence as a mutation affecting a single sample. Removing these sample mutations, the percentage of IS-consistent sites goes up from 40% to 92% (540,303 to 1,255,364 sites).

It would be interesting to remove all the mutations, remap them on using parsimony (map_mutations) and see if the 1,255,364 increases significantly.