FAQ

FAQ

My sample is Spoligo not found

The pattern is not in the Mbovis.org database. The Closest column lists the closest SB numbers (up to 3 spacers different). Either:

• It is a human-adapted lineage (M. tuberculosis): SB numbers only exist for the RD9-deleted lineages (M. bovis, M. caprae, M. pinnipedii, M. microti, M. africanum, ...). The Lineage column gives the lineage and its typical spoligotype families; use the octal code to find the shared international type (SIT) in the SITVIT database (H37Rv, 777777477760771, is SIT451).
• It is a new pattern: new M. bovis patterns can be submitted to Mbovis.org to get an SB number.
• A spacer is miscalled: look at the SpacerCount column for counts close to --min-count, and see How it works.

A sample with no spacer is SB2277

SB2277 is the pattern with no spacer at all. A sample that is not from the M. tuberculosis complex, or a file with no MTBC reads, also has no spacer, and so gets SB2277. spoligotyper flags these samples with a "no spacer found" warning, and the Species column says "MTBC not detected" when there is no MTBC DNA. M. canettii has no standard spacer either: it is reported with RD9 intact and no lineage.

How reliable are the species and the lineage?

They were checked on 16 reference genomes and 8 read sets of known species and lineage: see Validation. The lineage comes from the SNP barcode of Coll et al. (2014), the reference method for SNP-based lineage typing. For drug resistance and a finer lineage, use a dedicated tool such as TB-Profiler.

What does "only about X% of the reads appear to be from the M. tuberculosis complex" mean?

The MTBC-specific control regions have fewer reads than the sequencing depth predicts: part of the reads come from something else (contamination, host DNA, another organism). The spoligotype and species are still called from the MTBC reads. See Species and lineage.

Can I use another database?

Yes, with --db my_database.txt. The file has one pattern per line, with 3 columns separated by spaces or tabs: the octal code, the name, and the binary pattern. Lines starting with # are ignored. For example, with SIT numbers:

# octal          name    binary
777777477760771  SIT451  1111111111111111111001111111111100001111111

spoligotyper checks that each octal code matches its binary pattern.

How recent is the included database?

It is a snapshot of the Mbovis.org database with 1,976 SB patterns. To use a newer version, download it from Mbovis.org, format it as above and use --db.

Can I type nanopore reads?

Type the assembly, not the reads. In our hands, spoligotypes called directly from nanopore reads were often wrong, while spoligotypes called from assemblies of the same long reads were almost always right.

Spacers are 25 bp long and spoligotyper allows 1 mismatch per spacer: individual nanopore reads carry enough errors (substitutions and, above all, small insertions and deletions) that many reads covering a present spacer are missed, and present spacers can fall below --min-count. The consensus sequence of an assembly corrects these errors. So:

1. assemble the long reads (e.g. Flye or Autocycler), ideally with polishing (e.g. Medaka);
2. type the assembly: spoligotyper -r1 assembly.fasta -o results/.

The direct repeat locus is repetitive, but long reads usually span it entirely, so it assembles well. If you type nanopore reads anyway, treat the result as provisional: check the SpacerCount column for present spacers with low counts, and confirm with the assembly.

Does spoligotyper work on M. tuberculosis?

Yes: spacer detection and the binary, octal and hexadecimal codes work for the whole complex. Only the SB number is specific to the animal-adapted lineages (see above).

Why are there 43 spacers?

The 43 spacers are those of the standard spoligotyping membrane (Kamerbeek et al. 1997), so that in silico results can be compared with laboratory spoligotyping. Other spacers exist in some strains but are not part of the standard pattern.

How do I type many samples?

Put them in a folder and use -i: see Usage. fastq and fasta files are detected, and R1/R2 files are paired automatically. You get one table and one PDF report for the whole run.

Can the PDF report be used for accredited (ISO 17025) work?

It is designed for it: it records the operator, date and time with time zone, computer, exact command, parameters, versions of all the software, checksums of the input files and of the reference data, and the evidence behind each call, and it has a review and signature box. Validating the method for your scope remains your laboratory's responsibility; the tutorial data, with known spoligotypes, can be part of it.

Should I type reads or the assembly?

Reads, when you have them: the DR locus is repetitive and can be broken or collapsed in short-read assemblies. See How it works.

How do I cite spoligotyper?

See Home.