List all solid k-mers with DSK

Question

[alexandre.drouin.8]

Hi,

I am using DSK to count the k-mers that are present in multiple fasta files. I would like to get the list of all solid k-mers that were found in all files.

The command line interface has a "-solid-kmers-out" argument which, I assume, should output this list in a file. However, even when I set a value, it does not seem to output a file. Am I using it correctly?

Also, is there a programmatic way to go from the integer representation of k-mers to their string representation?

Thanks for your help, Alex

Answer 1

[Rayan Chikhi]

Bonjour Alexandre,

To output the solid kmers in string representation, you can either use the dsk2ascii tool provided with DSK, or parse the hdf5 file using h5dump. See https://github.com/GATB/dsk#results-visualization

To integrate DSK results in a C++ program, it is currently recommended to use the GATB-core library. Have a look at the doxygen page and also the example code in the GATB-core repository that demonstrate some custom processing:

• kmer10.cpp is the most up to date version of that snippet.
• kmer16.cpp shows how to get separate kmer counts (one abundance per file) over multiple files (multi-bank kmer counting)
• kmer11-15 are a bit more advanced

Thanks for pointing this out by the way, the "-solid-kmers-out" argument was actually made for GATB-core graph construction (to output the graph and the solid kmers in different files), and is not used in DSK. I've added a note for the future versions of DSK.

Answer 2

[alexandre.drouin.8]

Merci Rayan! I will have a look at the snippets.

Alex