LisSero

In silico serogroup typing prediction for Listeria monocytogenes

Authors

• Jason Kwong (@kwongjc) - GitHub: kwongj
• Josh Zhang (@abcdtree) - GitHub: abcdtree
• Torsten Seemann (@torstenseemann) - GitHub: tseemann

Maintainers

• MDU PHL - Josh Zhang, Kristy Horan, and Anders Gonçalves da Silva

Dependencies

• Python 3.6+
• BLAST 2.10.0

Installation

pip3 install lissero

Installing BLAST 2.10.0

Brew:

brew install blast

Conda:

conda install -c bioconda blast 

In silico serogrouping for Listeria monocytogenes

LisSero is based on a method of predicting serogroup for Listeria monocytogenes using PCR, as described by Doumith et al (see References).

It detects the presence or absence of 5 genes (lmo1118, lmo0737, ORF2110, ORF2819 and Prs).

The patterns obtained reflect the four main serotypes (1/2a, 1/2b, 1/2c, and 4b) obtained from food and human sources.

The patterns are not based on genes involved in somatic (O) or flagellar (H) biosynthesis, and are not specific to just one serotype, but rather to a group of serotypes.

If only Prs is detected, these isolates are often serotype 4a or 4c, though LisSero reports these as "Nontypable".

Usage

$ lissero --help

Usage: lissero [OPTIONS] FASTA...

  In silico serogroup prediction for L. monocytogenes. Alleles: lmo1118,
  lmo0737, ORF2819, ORF2110, Prs

  References:

  * Doumith et al. Differentiation of the major Listeria monocytogenes
  serovars by multiplex PCR. J Clin Microbiol, 2004; 42:8; 3819-22

Options:
  -s, --serotype_db TEXT
  --min_id FLOAT          Minimum percent identity to accept a match. [0-100]
                          [default: 95.0]

  --min_cov FLOAT         Minimum coverage of the gene to accept a match.
                          [0-100]  [default: 95.0]
  --logfile TEXT          Save log to a file instead of printing to stderr
  --debug
  --help    

Example usage

lissero my_fasta.fa

Example output

Output explained

The output consist of 8 columns:

• ID: At the moment the full path to your file
• SEROTYPE: A predicted serotype if possible, otherwise Nontypeable
• Five columns for each of the genes with either:
  • FULL: for a complete match to the gene in the database (as defined by the --min_cov and --min_id flags which default to at least 95% coverage and percent id)
  • NONE: when there is no match to the database
  • PARTIAL: where there are matches that fall below the --min_cov and --min_id thresholds
• COMMENTS: Which will try to explain a Nontypeable results. Current cases are:
  • No Prs found: when there is no FULL match to the Prs gene is not likely to be a Listeria monocytogenes
  • Presence of all 5 genes, not a known serotype: when all 5 genes are present
  • No combination of fully matched genes resulted in a known serotype: An unknown combination
  • In addition, it will say Unusual 4b with lmo0737 when reporting 4b, 4d, 4e* serotype

Change Log

Version 0.4.1

• Implemented unit test for all possible 32 possible gene patterns
• Implemented more verbose output with presence/absence info for each gene
• Implemented test of FASTA file, and will exit with warning if a file does not appear to be a FASTA file
• Implement test of whether an input file exists or not
• Implemented logging using Loguru
• Added option to log to a file (--logfile; stderr remains default)
• Logging now includes the BLAST command used
• Removed all references in the code to Binary Typing
• Implemented --version flag
• Added some better docs

Version 0.4.0

• Removed support for Binary Typing
• Consolidated repos in MDU-PHL
• Changed executable from run_lissero to lissero

Bugs

Please submit via the GitHub issues page: https://github.com/MDU-PHL/LisSero/issues

Software Licence

GPLv2: https://github.com/MDU-PHL/LisSero/blob/master/LICENSE

References

• Doumith et al. Differentiation of the major Listeria monocytogenes serovars by multiplex PCR. J Clin Microbiol, 2004; 42:8; 3819-22.