Buenos dias de la equipo del kmerdb 🇨🇴😄

[MatthewRalston]

K-mer database ?

Does the community need a tool that rapidly generates a representative k-mer signature, a spectrum, a positive-definite discreet column | for biology:

The k-mer array.

What would a k-mer data structure require? Mostly metadata:

• Where the k-mer was identified (typically a fasta id)
• What is the quantity of each k-mer observed across the sequence(s)?
• NOTE: Is the input .fa|.fq a properly qc'd sequence file?
• assuming the base-call is correct for the sequence file, what is the metadata about neighboring k-mers (the 6 nearest k-mer neighbors in sequence space [3 bases permuted on each side]

Possible questions for future work or expansions:

• How do we respect the graphical nature of the problem, and not reduce the problem to a compressed archive for primarily indexes on a single count column?
• If Neo4j or other graphical databases come into maturity and affordability, can we retool this to facilitate assemblies within those platforms, through their graph-solving capabilities?
• Can we also store other metadata linkages between k-mer nodes, for future explorations?
• How can we incorporate a mutational model into a k-mer array, to provide representative k-mer profiles derived from a parent model?

Welcome to the center of discussion behind a pointless, unfunded waste of computational resources that is the k-merdb project. Here we have a strict code of insults, social maladies, pathologies, and degeneracy to reinforce the hegemony that is the permissive license attached to this code: fork it if you don't want it. Still here? Please, be respectful at a minimum, civil at worst, and politely take your extra minute to describe what your interest is in the codebase, its future, or the technical topic at hand.