@Ammar Aziz right now the default functioning of kraken2 in viralrecon is set through these params:

kraken2_db                   = '<s3://ngi-igenomes/test-data/viralrecon/kraken2_human.tar.gz>'
    kraken2_db_name              = 'human'
    kraken2_variants_host_filter = false
    kraken2_assembly_host_filter = true

This means that: • For mapping+variant calling it WON'T remove host reads, just perform kraken2 to calculate host reads statistics, for example to know how well or bad amplicon primer enrichment worked. You can change this behaviour by changing

kraken2_variants_host_filter = false

to

true

• For de novo assembly, it WILL remove host reads prior to assembly (keeping anything that is unclassified) • The default host database is human, but by providing another kraken2 database through

kraken2_db

, you can use anyother host genome. You can even create a multiple host genomes database. For arboviruses we use a mixture of mosquitoes and vero (African Green monkey) genomes, as they are extracted from mosquitoes and then cultivated, and we find both genomes in the sequencing reads. Hope this solves your doubt!