MRG: add capacity for skipmer sketching and search

[bluegenes]

Skipmers are something we've considered adding for quite some time, as DNA kmer that ~allows "mismatches" aka increases entropy + sensitivity.

Over in sourmash-bio/sourmash#3395, I added a skipm1n3 and skipm2n3 moltypes, as well as code in SeqToHashes to build them. In sourmash-bio/sourmash#3446 I also added capacity for sourmash python function to read skipmer sigs, so sig cat, sig summarize, etc should now work.

There are two types of skipmers available, keep-2,skip-1 ("skipm2n3") and keep-1,skip-2 ("skipm1n3"). To sketch with skipmers, specify skipm2n3 or skipm1n3 in the parameter string. The skipmer ksize is the "final" size that the k-mer ends up. --i.e. for ksize 3, the sequence ACTAG would produce two skip-mers for m2n3: ACA, CTG.

example sketching commands:

manysketch:

sourmash scripts manysketch -p skipm2n3,k=21,scaled=1000 ms.csv -o output.zip

singlesketch:

sourmash scripts singlesketch -p skipm2n3,k=21,scaled=1000 myfile.fasta -o myfile.sig.gz

-o myfile.zip also works

Skipmer References:

• Skip-mers: increasing entropy and sensitivity to detect conserved genic regions with simple cyclic q-grams
• Extracting and Evaluating Features from RNA Virus Sequences to Predict Host Species Susceptibility Using Deep Learning

ref #549

[bluegenes]

prior to merge (or at least release), need to switch to sourmash 0.18.0 in Cargo.toml