I am currently working on developing a pipeline to optimize k-mer counting. I have some ideas, however, they would work only for small k-values. I am trying to figure out whether k-mer counting is used with small k-values, especially k
Small kmers are used by several MinHash based programs that seek to very rapidly give you distance estimates between 2 sequences (or all pairwise distances from a collection of sequences). Mash is the one I most commonly use, https://mash.readthedocs.io/en/latest/ . The default kmer size it uses is 21 though.