Allow specification of population prior allele frequencies with mchap call

[timothymillar]

There are a number of components to consider:

• [ ] Input of prior frequencies in VCF file

  • Use of allele counts field?
  • Use of an optional field to normalize?
  • Parameter to flatten the input distribution?

• [ ] Integration of prior frequencies

  • Needs to be added to likelihood function so will need to passed through a lot of jitted code

• [ ] Reporting of prior frequencies?

  • Use of an INFO field of length alleles
  • This affects the posterior distribution so best to (optionally) report it for transparency
  • May also allow for reusing the same prior in another analysis

[timothymillar]

This also raises the possibility of reporting the population posterior distribution of alleles in both assemble and call tools. Therefore related to #103