Home

Awesome

PLEASE NOTE - THIS FORK IS PROVIDED AS-IS

This fork is provided as is, I'm not the maintainer, just a bioinformatician that wanted to use PMGA (PubMLST Genome Annotator v2.0) on the command-line. The folks at the CDC, in particular Nadav Topaz deserves all the kudos and credit for creating PMGA!

Given I only wanted to run this on the command-line, please keep that in mind if any issues arise. I might be able to assist with technical issues, but I cannot help interpret the results (e.g. serotype/serogroup). For help with the biological significance of the results, it would probably be best to reach out to Bacterial Meningitis Genome Analysis Platform (BMGAP).

I'm sure they would be willing to help!

To review the original docs I recommend you visit: PMGA (PubMLST Genome Annotator v2.0)

PMGA - PubMLST Genome Annotator

PMGA is a tool for serotyping/serogrouping all Neisseria species and Haemophilus influenzae. PMGA requires you to build a set of BLAST databases using data that is available from PubMLST. You can then query your uncompressed FASTA files against the BLAST databases. If you do not provide a species using the --species option, Mash will be used to determine the species. Once complete, you will get output files (discussed below).

Installation

Currently pmga is available on my personal Conda channel, but is expected to become available on BioConda soon.

mamba create -n pmga -c pmga -c conda-forge -c bioconda pmga

Building BLAST Databases

Prior to running pmga, you will need to build the BLAST databases with pmga-build. During this process, pmga-build will download each loci for all schemes associated with Neisseria scpecies and Haemophilus influenzae. Building the databases can take upwards to 90 minutes, but this is a one time step. After pmga-build is completed, by defualt all the outputs will be written to ./pubmlst_dbs_all.

pmga-build Usage

pmga-build --help
usage: pmga-build [-h] [--outdir STR] [--force] [--silent] [--debug] [--version]

pmga-build - Script for creating local BlastDBs from PubMLST alleles

options:
  -h, --help    show this help message and exit
  --outdir STR  Directory to save BLAST databases to (Default: ./pubmlst_dbs_all)
  --force       Overwrite existing directories.
  --silent      Only critical errors will be printed.
  --debug       Print debug related text.
  --version     show program's version number and exit

Running pmga

After you have sucessfully build all the necessary BLAST databases, you are now ready to start serotyping/serogrouping your Neisseria and H. influenzae samples! pmga can be executed on an uncompressed FASTA file and will output a number of files.

pmga Usage

pmga --help
usage: pmga [-h] [--prefix STR] [--blastdir STR] [--species STR] [-t INT] [-o STR] [--force] [--verbose]
            [--silent] [--version]
            FASTA

pmga - Serotyping, serotyping and MLST of all Neisseria species and Haemophilus influenzae

positional arguments:
  FASTA                 Input FASTA file to analyze

options:
  -h, --help            show this help message and exit
  --prefix STR          Prefix for outputs (Default: Use basename of input FASTA file)
  --blastdir STR        Directory containing BLAST DBs built by pmga-build (Default: ./pubmlst_dbs_all

Additional Options:
  --species STR         Use this as the input species (Default: use Mash distance). Available Choices:
                        neisseria, hinfluenzae
  -t INT, --threads INT
                        Number of cores to use (default=1)
  -o STR, --outdir STR  Directory to output results to (Default: ./pmga)
  --force               Force overwrite existing output file
  --verbose             Print debug related text.
  --silent              Only critical errors will be printed.
  --version             show program's version number and exit

pmga Output Files

Below are the expected outputs from each pmga run. The files will be output to the value set by --outdir, which defaults to ./pmga. Each of the output files will use the value of --prefix for filenames. By default the prefix is set to the basename of the input file.

<OUTDIR>
├── <PREFIX>-allele-matrix.txt
├── <PREFIX>-final-blast-results.json.gz
├── <PREFIX>-loci-counts.txt
├── <PREFIX>-raw-blast-results.json.gz
├── <PREFIX>.gff.gz
└── <PREFIX>.txt
ExtensionDescription
*-allele-matrix.txtA tab-delimitted file with the allele ID for each loci with a hit
*-blast-final-results.json.gzFiltered BLAST results in JSON format
*-blast-raw-results.json.gzUnfiltered BLAST results in JSON format
*-loci-counts.txtA tab-delimitted file with the number of hits per loci
*.gff.gzA GFF3 file annotated with the BALST results
*.txtA tab-delimitted file with final predicted serotype/serogroup for the sample

Example Serotype/Serogroup Output

sample	species	prediction	genes_present	notes
GCF_003355215	neisseria_serogroup	B	csb,cssA,cssB,cssC,ctrA,ctrB,ctrC,ctrD,ctrE,ctrF,tex	B backbone: All essential capsule genes intact and present

Above an example of the output predictions, which contains 5 columns. The 5 columns are

ColumnDescription
sampleThe input sample name
speciesThe species of the sample either neisseria_serogroup or hinfluenzae_serotype
predictionThe predicted serotype or serogroup
genes_presentA list of genes present in the sample
notesAny notes associated with the prediction

Citations

If you make use of this tool, please cite the following: