Add function to read fasta subset

[zachary-foster]

Big fasta files are common. I just ran into one that would require 14GB of RAM to read into R using ape::read.dna. I need a way to read into subsets of large fasta files.

The following should be possible:

• Read in a defined set of sequence indexes (e.g. c(3, 2) would read the third and second sequence)
• Read a random subset. This can use the code for the defined subset, but requires knowing the number of sequences in the file.

[zachary-foster]

@knausb suggested using readr for handling gzipped file inputs