Exclude sites that are prone to sequencing or base-calling errors

[Haddox]

See Methods section of Bloom and Neher, in the subsection called "Counting mutations along the phylogenetic tree".

[Haddox]

@chrisjenningsshaffer: here are my thoughts on which sites we should mask. They include error-prone sites with data (counts might be inflated at those sites) and sites masked in the UShER tree (I believe that means the tree does not report counts at these sites, so we cannot estimate fitness effects of mutations at these sites).

First, we should mask the same set of sites masked in Jesse and Richard's original pipeline. See lines 81-113 of their config file for a list of sites to exclude as well as files giving lists of sites masked in the UShER tree. We should mask a site even if it's masked in only one clade. You could look at the scripts in the snakemake pipeline to see how Jesse reads the files with lists of sites to mask.

Second, we should also mask the following set of sites from De Maio et al.:
187, 1059, 2094, 3037, 3130, 6990, 8022, 10323, 10741, 11074, 13408,
14786, 19684, 20148, 21137, 24034, 24378, 25563, 26144, 26461, 26681, 28077,
28826, 28854, 29700, 4050, 13402, 11083, 15324, 21575

This should result in the same masking strategy that we're using in the synonymous-mutations project. Does that sound good to you?

[chrisjenningsshaffer]

Yes, I can work with that.