Enquiry: Kraken2 result. Pear sequences tagged as apple

[CeciliaDeng]

Description of the bug

I used the assemblyQC to check the quality of transcripts in several published Pear genomes. The BUSCO completeness scores look good (>95% complete completeness). In the Kraken result, all the pear transcripts were classified as apple. For example, BartDHv2 was classified 100% Malus, broken down to 41% M. domestica, 49% M. sylvestris. The remaining 10% was not labeled and the color looks like Malus. Why were the 10% transcripts not labelled to species level? And why the whole transcriptome was classified as apple, not pear (Pyrus)? Was Pyrus not included in the Kraken2 DB? Thank you.

Command used and terminal output

No response

Relevant files

No response

System information

No response

[GallVp]

At PFR, we are using k2_pluspfp_20230314 database. It does include plant data.

A new update has been released here: https://benlangmead.github.io/aws-indexes/k2

Updating the data might fix this issue. Not sure, though. Kraken2 is not the most accurate tool.

[GallVp]

See the database update issue #170

[GallVp]

Hi @CeciliaDeng

I have updated the Kraken 2 database. Can you please run your data with the new version 2.2.0 and see if it has resolved your problem. Thank you!

[GallVp]

From MS Teams,

Results for pear genomes (and transcriptomes) remain the same with the update kraken2 DB.