A tool easily taken advantage of for in silico serogrouping of Pseudomonas aeruginosa isolates
Pseudomonas aeruginosa serotyper (PAst) is available as a web-service from CGE - PAst. The web-service works great, but is not the most efficient for 100s of genomes. I also looked into the original implementation available at PAst. I tried my best to start tweaking it to meet my needs, but unfortunately my Perl skills have dulled over the years! And it ended up being easier and quicker to just rewrite it in Python.
pasty is built using camlhmp to identify the
serogroup of Pseudomonas aeruginosa isolates. Using an input assembly (uncompressed or
gzip-compressed), the sequences are blasted against a set of O-antigens. The serogroup is
then predicted based on these results.
The serogroup logic is based on Table 1 of the original PAst publication (citation below).
pasty Is available from Bioconda
mamba create -n pasty -c conda-forge -c bioconda pasty
conda activate pasty
pasty --help
Usage: camlhmp-blast-region [OPTIONS]
🐪 camlhmp-blast-region 🐪 - Classify assemblies with a camlhmp schema using BLAST against
larger genomic regions
╭─ Options ─────────────────────────────────────────────────────────────────────────────────────╮
│ * --input -i TEXT Input file in FASTA format to classify [required] │
│ * --yaml -y TEXT YAML file documenting the targets and types │
│ [default: bin/../data/pa-osa.yaml] │
│ [required] │
│ * --targets -t TEXT Query targets in FASTA format │
│ [default: bin/../data/pa-osa.fasta] │
│ [required] │
│ --outdir -o PATH Directory to write output [default: ./] │
│ --prefix -p TEXT Prefix to use for output files [default: camlhmp] │
│ --min-pident INTEGER Minimum percent identity to count a hit [default: 95] │
│ --min-coverage INTEGER Minimum percent coverage to count a hit [default: 95] │
│ --force Overwrite existing reports │
│ --verbose Increase the verbosity of output │
│ --silent Only critical errors will be printed │
│ --version -V Print schema and camlhmp version │
│ --help Show this message and exit. │
╰───────────────────────────────────────────────────────────────────────────────────────────────╯
An assembly in FASTA format (compressed with gzip, or uncompressed) to predict the serogroup on.
Note: These should be automatically set for you!
These are the required inputs for camlhmp to specify the schema and targets for
serogrouping P. aeruginosa.
The prefix to use for the output files. If a prefix is not given, the basename of the input assembly will be used.
The directory to save result files. If the directory does not exist is will be created.
The minimum percent identity for a blast hit to be considered for serogrouping. The is compared
against the pident column of the blast output.
The minimum coverage of a O-antigen to be considered for serogrouping. This looks at the percent
of the O-antigen that was aligned to. This calculation does include mismatches and gaps,
but those should be expected to be minimal if --min_pident is set to the default (95%) or
similar.
| Filename | Description |
|---|---|
{PREFIX}.tsv |
A tab-delimited file with the predicted serogroup |
{PREFIX}.blastn.tsv |
A tab-delimited file of all blast hits |
{PREFIX}.details.tsv |
A tab-delimited file with details for each serogroup |
This file will contain the final predicted serogroup based on highest coverage. Here's what to expect the output to look like:
sample type targets coverage hits schema schema_version camlhmp_version params comment
O1-GCF_001420225 O1 O1 100.00 1 pasty 1.0.0 0.2.1 min-coverage=95;min-pident=95
| Column Name | Description |
|---|---|
| sample | Name of the sample processed |
| type | The predicted serogroup |
| targets | The targets for the serogroup (type column) with a passing hit |
| coverage | The percent of the O-antigen that was aligned to |
| hits | The number of blast hits included in the prediction (fewer the better) |
| schema | The schema used for the prediction |
| schema_version | The version of the schema used for the prediction |
| camlhmp_version | The version of camlhmp used for the prediction |
| params | The parameters used for the prediction |
| comment | Any comments produced during analysis |
The blast results are in standard -outfmt 6, and will be similar to this:
qseqid sseqid pident qcovs qlen slen length nident mismatch gapopen qstart qend sstart send evalue bitscore
O1 NZ_LJNU01000023.1 100.000 100 18368 73203 18368 18368 0 0 1 18368 21280 39647 0.0 33920
O2 NZ_LJNU01000023.1 97.823 10 23303 73203 1975 1932 41 2 1 1974 39647 37674 0.0 3408
O2 NZ_LJNU01000023.1 96.296 10 23303 73203 324 312 11 1 22980 23303 21602 21280 2.06e-149 531
O3 NZ_LJNU01000023.1 97.468 11 20210 73203 1975 1925 48 2 1 1974 39647 37674 0.0 3369
O3 NZ_LJNU01000023.1 100.000 11 20210 73203 292 292 0 0 19919 20210 21571 21280 2.96e-152 540
O4 NZ_LJNU01000023.1 96.350 14 15279 73203 1918 1848 70 0 1 1918 39647 37730 0.0 3155
O4 NZ_LJNU01000023.1 99.275 14 15279 73203 276 274 2 0 15004 15279 21555 21280 3.80e-140 499
O6 NZ_LJNU01000023.1 97.809 14 15649 73203 1917 1875 42 0 1 1917 39647 37731 0.0 3308
O6 NZ_LJNU01000023.1 99.649 14 15649 73203 285 284 1 0 15365 15649 21564 21280 8.31e-147 521
O7 NZ_LJNU01000023.1 97.722 12 19617 73203 1975 1930 43 2 17644 19617 37674 39647 0.0 3397
O7 NZ_LJNU01000023.1 100.000 12 19617 73203 292 292 0 0 1 292 21280 21571 2.88e-152 540
O9 NZ_LJNU01000023.1 98.010 39 17263 73203 6332 6206 122 2 1 6331 39647 33319 0.0 10992
O9 NZ_LJNU01000023.1 88.664 39 17263 73203 494 438 39 12 16778 17263 21764 21280 3.20e-166 586
O10 NZ_LJNU01000023.1 98.382 13 17635 73203 1916 1885 31 0 15720 17635 37732 39647 0.0 3367
O10 NZ_LJNU01000023.1 98.973 13 17635 73203 292 289 3 0 1 292 21280 21571 2.60e-147 523
O11 NZ_LJNU01000023.1 96.296 16 13868 73203 1917 1846 71 0 11952 13868 37731 39647 0.0 3147
O11 NZ_LJNU01000023.1 98.932 16 13868 73203 281 278 3 0 1 281 21280 21560 2.67e-141 503
O12 NZ_LJNU01000023.1 90.449 9 25864 73203 1916 1733 183 0 23949 25864 37732 39647 0.0 2525
O12 NZ_LJNU01000023.1 95.455 9 25864 73203 330 315 11 3 1 328 21280 21607 3.82e-147 523
O13 NZ_LJNU01000023.1 97.767 15 14316 73203 1926 1883 43 0 1 1926 39647 37722 0.0 3319
O13 NZ_LJNU01000023.1 99.650 15 14316 73203 286 285 1 0 14031 14316 21565 21280 2.11e-147 523
This file provides the the coverage and number of hits for each of the serogroups. This can be useful for a deeper review, and it should look like the following:
sample type status targets missing coverage hits schema schema_version camlhmp_version params comment
O1-GCF_001420225 O1 True O1 100.00 1 pasty 1.0.0 0.2.1 min-coverage=95;min-pident=95
O1-GCF_001420225 O2 False O2,wzyB 9.86,0.00 2,0 pasty 1.0.0 0.2.1 min-coverage=95;min-pident=95 O2:Coverage based on 2 hits
O1-GCF_001420225 O3 False O3 11.21 2 pasty 1.0.0 0.2.1 min-coverage=95;min-pident=95 Coverage based on 2 hits
O1-GCF_001420225 O4 False O4 14.36 2 pasty 1.0.0 0.2.1 min-coverage=95;min-pident=95 Coverage based on 2 hits
O1-GCF_001420225 O5 False O2 9.86 2 pasty 1.0.0 0.2.1 min-coverage=95;min-pident=95 Coverage based on 2 hits
O1-GCF_001420225 O6 False O6 14.07 2 pasty 1.0.0 0.2.1 min-coverage=95;min-pident=95 Coverage based on 2 hits
O1-GCF_001420225 O7 False O7 11.55 2 pasty 1.0.0 0.2.1 min-coverage=95;min-pident=95 Coverage based on 2 hits
O1-GCF_001420225 O9 False O9 36.67 1 pasty 1.0.0 0.2.1 min-coverage=95;min-pident=95
O1-GCF_001420225 O10 False O10 12.52 2 pasty 1.0.0 0.2.1 min-coverage=95;min-pident=95 Coverage based on 2 hits
O1-GCF_001420225 O11 False O11 15.85 2 pasty 1.0.0 0.2.1 min-coverage=95;min-pident=95 Coverage based on 2 hits
O1-GCF_001420225 O12 False O12 1.27 1 pasty 1.0.0 0.2.1 min-coverage=95;min-pident=95
O1-GCF_001420225 O13 False O13 15.45 2 pasty 1.0.0 0.2.1 min-coverage=95;min-pident=95 Coverage based on 2 hits
| Column Name | Description |
|---|---|
| sample | Name of the sample processed |
| type | The predicted serogroup |
| targets | The targets for the serogroup (type column) with a passing hit |
| missing | The targets for the serogroup (type column) without a passing hit |
| coverage | The percent of the O-antigen that was aligned to |
| hits | The number of blast hits included in the prediction (fewer the better) |
| schema | The schema used for the prediction |
| schema_version | The version of the schema used for the prediction |
| camlhmp_version | The version of camlhmp used for the prediction |
| params | The parameters used for the prediction |
| comment | Any comments produced during analysis |
The original implementation was called PAst, and given this is a Python version I originally
went with PAst-py. But decided, "nah, let's go with pasty". Hopefully pasty will be
your patsy when it comes to serogrouping your P. aeruginosa isolates.
The original PAst did not include a license, so I selected Apache 2.0 as the license to match other CGE tools.
If you use this tool please cite the following:
Serogroup Method
Thrane SW, Taylor VL, Lund O, Lam JS, Jelsbak L Application of Whole-Genome Sequencing Data for O-Specific Antigen Analysis and In Silico Serotyping of Pseudomonas aeruginosa Isolates. Journal of Clinical Microbiology, 54(7), 1782–1788. (2016)
pasty
Petit III RA pasty: A tool easily taken advantage of for in silico serogrouping of Pseudomonas aeruginosa isolates (GitHub)
camlhmp
Petit III RA camlhmp: Classification through yAML Heuristic Mapping Protocol (GitHub)
BLAST+
Camacho C, Coulouris G, Avagyan V, Ma N, Papadopoulos J, Bealer K, Madden TL BLAST+: architecture and applications. BMC Bioinformatics 10, 421 (2009)