Random genration of Kmers

Kmers.jl supports efficient random generation of kmers through a package extension, when the Random stdlib is loaded.

Sampling a symbol from a kmer

julia> m = mer"PLQWHY"a;

julia> rand(m)
AA_L

julia> rand(m)
AA_H

When sampling a random kmer from a type, kmers are sampled from the following distributions:

For T <: AAKmer, the kmer symbols is uniformly sampled from the 20 standard proteogenic amino acids ACDEFGHIKLMNPQRSTVWY. This excludes amino acids such as selenocysteine, and pseudo-aas such as AA_Term and AA_Gap
For T <: DNA{<:NucleicAcidAlphabet{4}}, the kmer symbols is uniformly sampled from the 4 unambiguous nucleotides. E.g. for T <: Kmer{DNAAlphabet{4}}, the only symbols are A, C, G and T.
For all other kmer types, the symbols are sampled uniformly from all symbols in its alphabet.

You can sample from either a concrete type, or a type with the trailing N type parameter elided:

julia> rand(DNAKmer{9})
DNA 9-mer:
TGGGAGCCA

julia> rand(AAKmer{4, 1})
AminoAcid 4-mer:
RRIP