implementation of k-mer counting from krakenuniq to kraken2

Question

[aziznasr1920]

Hello, I read that Kraken2 support now k-mer counting from KrakenUniq. Can anyone help me in the implementation of this feature on Kraken2.

Thanks in advance. Have a good day!

[acvill]

"I read that Kraken2 support now k-mer counting" -- where did you read this?

[aziznasr1920]

acvill , GenoMax Thank you for your responses. This is the source where I read that kraken2 support k-mer counting of krakenuniq: https://github.com/DerrickWood/kraken2/pull/220
I remember that I read a paper talking about that also.

For the databases problem, I will not run krakenuniq on kraken2 database, I will just implement a feature of rakenuniq in kraken2, but the tool still kraken2.

[GenoMax]

I will just implement a feature of rakenuniq in kraken2, but the tool still kraken2.

Based on this comment from the thread you linked, this implementation has been already done. You should see the results in kraken2 report.

Answer 1

[h.mon]

Old post, but the relevant information has been updated and published.

One can use the flag --report-minimizer-data to get the unique k-mer counts per classification. See Kraken2 page, or the protocol "Metagenome analysis using the Kraken software suite".

Answer 2

[GenoMax]

Unless I am reading this wrong, authors say that

KrakenUniq is only compatible with Kraken 1 databases, not Kraken 2.

You may be able to choose a classifier using the advice on that page.