I/O to import genotype data from standard file formats

[quattro]

Catch-all for implementing I/O routines to load genotype data from standardized data formats. Namely,

• PLINK formatted data
• PLINK2 formatted data
• VCF (gzip VCF) formatted data
• BGEN formatted data

We can leverage existing libraries as dependencies to load in these data when possible, but might require some custom code to avoid materializing entire dense genotype vectors/matrices in memory.

[quattro]

Let's use some of the sparse readers implemented here: https://github.com/fastlmm/bed-reader