Example of Kraken2 report


Kraken2 is used to classify reads according to their most likely taxonomic clade.

More: speciation documentation

Kraken2 DB: MiniKraken2_v2_8GB

See how kraken output is parsed.

Mykrobe Speciation

Example of Mykrobe report


Mykrobe is used to further classification if the sample appear to be mycobacterial, or a mixture containing M. tuberculosis reads.

Mykrobe tests for the presence of probe sequences in the reads to determine the species present.

The “Coverage” is the percent of the probe that has non-zero read depth. “Median depth” is the median read depth across the probe.

See mykrobe documentation

See SNP panel for lineage classification

See how mykrobe output is parsed.


Example of Reference report


Reference is selected based on mykrobe result.

There are currently 10 available, with associated RefSeq chromosome IDs as follows:

  • Mycobacterium abscessus –> NC_010397.1

  • Mycobacterium africanum –> NC_000962.3

  • Mycobacterium avium –> NC_002944.2

  • Mycobacterium bovis –> NC_002945.4

  • Mycobacterium chelonae –> NZ_CP007220.1

  • Mycobacterium chimaera –> NZ_CP012885.2

  • Mycobacterium fortuitum –> NZ_CP011269.1

  • Mycobacterium intracellulare –> NC_016946.1

  • Mycobacterium kansasii –> NC_022663.1

  • Mycobacterium tuberculosis –> NC_000962.3

Samtools QC

Example of QC report


See Samtools documentation


Example of drug resistance report


Proceed only if M. tuberculosus:

  • First a Genome object of the M. tuberculosis reference (NC_000962.3) is created/loaded (reference) using gumpy

  • It is then copied for the sample

  • The per-sample VCF produced by Clockwork is consumed and applied to the sample Genome object.

  • gumpy methods then produce lists of the genetic variants and mutations between reference and sample.

  • These are then parsed by the piezo package which in turn requires a resistance catalogue stored using a standard grammar (GARC). Some published catalogues are available here (others will be added over time, including the Seq&Treat/WHO catalogue and any resulting from CRyPTIC).

Resistance configuration

  • The antibiogram is then output for the drugs contained in the specified catalogue.

  • Also calculated and stored, but not shown in the report at present, are the effects (R/S/U) on each drug of each relevant mutation, allowing one to, for example, understand what mutation was responsible for a specific drug being predicted to be resistant.


Which drugs are considered depends on which tuberculosis AMR catalogue was specified.


gumpy produces two hierarchies of genetic variant. The lower-level description contains all insertions/deletions and all single nucleotide polymorphisms (SNP) – we call these genetic variants. The higher-level description (which is typically used in resistance prediction) translates the latter into amino acids, if they lie within the coding region of a protein gene, otherwise there is no change. We describe these as mutations. Because more than SNP can occur in a codon, the latter tends to


Example of SNP distance report


SNP distance is calculated for a sample against other samples in the same run.