Troubleshooting

Troubleshooting

"seal.sh" was not found

spoligotyper uses the seal.sh found in your PATH first, then the one installed next to its own Python interpreter (the bin/ folder of the conda environment). Calling /path/to/envs/spoligotyper/bin/spoligotyper without activating the environment therefore works if BBTools is installed in that environment.

Otherwise, install BBTools: conda install -c bioconda bbmap.

Seal failed

The message ends with the last lines printed by Seal. Common causes:

• Java errors (java: command not found, UnsupportedClassVersionError): BBTools needs Java. Installing BBTools with conda installs a compatible Java.
• Truncated or corrupted input file: check it with gzip -t file.fastq.gz.

Run with -v to see the full Seal command and output.

Java memory errors

spoligotyper gives Seal 1 GB of memory (--memory 1g), which is plenty for 43 spacers: Seal's memory does not grow with the size of the input. If Java cannot reserve it (Could not reserve enough space for object heap, e.g. on a login node or in a small container), lower it: --memory 500m.

Before version 0.2.0, Seal sized its memory from the computer's free memory (196 GB on a 512 GB server), which failed on shared computers and clusters.

Several files give the same sample name

In batch mode (-i), each sample name must come from a single fasta file, a single fastq file, or an R1/R2 pair. The message lists the conflicting files: rename, move or remove them. See Usage for how sample names are derived from file names.

Some samples failed

In batch mode, a sample that cannot be typed (truncated file, not a sequence file, ...) is reported as failed with its error in the table and the PDF report, and the other samples are still typed. spoligotyper exits with code 1.

mixed sample?

The lineage SNPs, a region of difference or the spacer counts suggest several strains. See Species and lineage. Check the sample (e.g. single colony re-culture) before reporting its spoligotype: the pattern of a mixed sample is the union of those of its strains.

too little MTBC DNA to check the species

Spacers were found but the MTBC control regions have fewer than 3 reads each: the depth is too low, or most reads come from something else. The species and lineage are not called.

No spacer found

None of the 43 spacers is in the file. Check that the sample is from the M. tuberculosis complex (e.g. with Kraken2 or mashID), and that the file is not empty or truncated.

spacer(s) called absent were seen in fewer than 5 reads

Some spacers were found, but in too few reads to be called present. This happens with low coverage data, or with mixed or contaminated samples. See How it works to decide whether to lower --min-count.

assembly typed with minimum count N: spacers are probably missed

Assemblies contain each spacer once. Do not set --min-count for assemblies: it is 1 by default.

typed as reads in fasta format

The fasta file has more than 1,000 sequences and more than 13 Mb (3 times the genome): it holds reads (e.g. from fasterq-dump --fasta), not an assembly, and is typed as reads (minimum count 5, depth and MTBC fraction estimated). If it really is an assembly (e.g. a metagenome assembly), use --min-count 1.

Paths with spaces, commas, "xmx" or "xms"

They are supported. Seal cannot read them (BBTools splits arguments on spaces and commas, and reads any argument containing "xmx" or "xms" as a Java memory setting), so spoligotyper gives Seal links with neutral names in a temporary folder. The same links give Seal the right extension for files named without one (e.g. Galaxy's .dat files) or gzipped without .gz. Only the system temporary folder itself must be free of these (set TMPDIR otherwise).

-r2 is only for paired-end fastq files

-r2 must be the R2 reads of a paired-end run. Assemblies are given with -r1 alone.

ModuleNotFoundError: No module named 'pkg_resources'

That is version 0.1, which does not work with recent versions of setuptools. Update to version 0.2.0 or later (see Installation).