If kraken*report contains samples with a line containing "unclassfied" reads and other samples containing no such line then multiqc will interpret the former as "kraken" and the latter as "bracken" output - this will result in two different sections and also screw the parsing of the multiqc_report.json
An example is here: 20230607_0958_P2-7085206e51685c-B_PAO92085_9bf914e7
If kraken*report contains samples with a line containing "unclassfied" reads and other samples containing no such line then multiqc will interpret the former as "kraken" and the latter as "bracken" output - this will result in two different sections and also screw the parsing of the multiqc_report.json
An example is here: 20230607_0958_P2-7085206e51685c-B_PAO92085_9bf914e7