Trouble visualising kraken2 output

[archanamadhav]

Great tool!

Can you please tell me why the Sankey output looks like this? I have used Pavian before and have not encountered this previously.

[gregorysprenger]

Can you share your Kraken2 command?

[fbreitwieser]

Hi @archanamadhav , that's seems like a weird behavior of the Sankey tool. Can you share the kraken2 report file?

[gregorysprenger]

@fbreitwieser If the kraken2 param --report-zero-counts is added, it can produce a similar image. A simple awk command that filters out the zeros in the second column fixes the issue.

[archanamadhav]

Hi @fbreitwieser and @gregorysprenger!

Thank you for your responses! :)

Here is my Kraken2 command:

while read file; do kraken2 --db /rds/project/rds-Jz6YwN89pIk/kraken2_db/kraken2_new/custom_plusPF --paired kneaddata_output/${file%%.*}_1_kneaddata_paired_1.fastq.gz kneaddata_output/${file%%.*}_1_kneaddata_paired_2.fastq.gz --threads 50 --gzip-compressed --use-names --report kraken_reports/${file%%.*}.kreport2 --output kraken_reports/${file%%.*}.kraken2; done<list_of_ids.txt

I am also attaching a couple of example output files from my run. If its useful, these are enriched metagenome samples - so perhaps it could be related to that? I have not encountered this when using Pavian for "true" metagenomes. I should also point out that this happened with most but not all the report files generated in the same run.

For example: File 1 shows proper segmentation for E.coli but File 2 does not.

File1: NG-32851_P21BT0071_lib681543_10216_2.txt
File2: NG-32851_P21AT0584_lib681546_10222_1.txt