run examples/genome_scan.py on new data

[andrewkern]

Test the genome scan workflow on a new dataset beyond the bundled Ag1000G subset.

• Identify a suitable publicly available dataset (e.g. 1000 Genomes, DPGP, etc.)
• Run examples/genome_scan.py end-to-end with both VCF and zarr input
• Verify population assignment and two-population statistics work correctly
• Confirm figure output looks reasonable

[mufernando]

I could run this on great apes data and compare to my estimates using scikit-allel. the only problem is I don't have runtimes for the scikit-allel examples.

[andrewkern]

don't worry about scikit-allel timing -- I have that covered. It would be great to just run this and see how it goes.