This workflow takes provided JSON-formatted MLST profiles and converts them into a phylogenetic tree with associated flat cluster codes for use in Irida Next. The workflow also generates an interactive tree for visualization.
A brief overview of the usage of this pipeline is given below. Detailed documentation can be found in the docs/ directory.
The input to the pipeline is a standard sample sheet (passed as --input samplesheet.csv) that looks like:
| sample | mlst_alleles | metadata_1 | metadata_2 | metadata_3 | metadata_4 | metadata_5 | metadata_6 | metadata_7 | metadata_8 |
|---|---|---|---|---|---|---|---|---|---|
| SampleA | sampleA.mlst.json | meta1 | meta2 | meta3 | meta4 | meta5 | meta6 | meta7 | meta8 |
The structure of this file is defined in assets/schema_input.json. Validation of the sample sheet is performed by nf-validation. Details on the columns can be found in the Full samplesheet documentation.
gasclustering accepts the IRIDA-Next format for samplesheets which can contain an additional column: sample_name
sample_name: An optional column, that overrides sample for outputs (filenames and sample names) and reference assembly identification.
sample_name allows more flexibility in naming output files or sample identification. Unlike sample, sample_name is not required to contain unique values. Nextflow requires unique sample names, and therefore in the instance of repeat sample_names, sample will be suffixed to any sample_name. Non-alphanumeric characters (excluding _,-,.) will be replaced with "_".
An example samplesheet has been provided with the pipeline.
The main parameters are --input as defined above and --output for specifying the output results directory. You may wish to provide -profile singularity to specify the use of singularity containers and -r [branch] to specify which GitHub branch you would like to run.
In order to customize metadata headers, the parameters --metadata_1_header through --metadata_8_header may be specified. These parameters are used to re-name the headers in the final metadata table from the defaults (e.g., rename metadata_1 to country).
The Genomic Address Service Clustering workflow can use two distance methods: Hamming or scaled.
Hamming distances are integers representing the number of differing loci between two sequences and will range between [0, n], where n is the total number of loci. When using Hamming distances, you must specify --pd_distm hamming and provide Hamming distance thresholds as integers between [0, n]: --gm_thresholds "10,5,0" (10, 5, and 0 loci).
Scaled distances are floats representing the percentage of differing loci between two sequences and will range between [0.0, 100.0]. When using scaled distances, you must specify --pd_distm scaled and provide percentages between [0.0, 100.0] as thresholds: --gm_thresholds "50,20,0" (50%, 20%, and 0% of loci).
The --gm_thresholds parameter is used to set thresholds for each cluster level, which in turn are used to assign cluster codes at each level. When specifying --pd_distm hamming and --gm_thresholds "10,5,0", all sequences that have no more than 10 loci differences will be assigned the same cluster code for the first level, no more than 5 for the second level, and only sequences that have no loci differences will be assigned the same cluster code for the third level.
The following can be used to adjust parameters for the profile_dists tool.
--pd_outfmt: The output format for distances. For this pipeline the only valid value is matrix (required by gas mcluster).--pd_distm: The distance method/unit, either hamming or scaled. For hamming distances, the distance values will be a non-negative integer. For scaled distances, the distance values are between 0.0 and 100.0. Please see the Distance Method and Thresholds section for more information.--pd_missing_threshold: The maximum proportion of missing data per locus for a locus to be kept in the analysis. Values from 0 to 1.--pd_sample_quality_threshold: The maximum proportion of missing data per sample for a sample to be kept in the analysis. Values from 0 to 1.--pd_file_type: Output format file type. One of text or parquet.--pd_mapping_file: A file used to map allele codes to integers for internal distance calculations. This is the same file as produced from the profile dists step (the allele_map.json file). Normally, this is unneeded unless you wish to override the automated process of mapping alleles to integers.--pd_skip: Skip QA/QC steps. Can be used as a flag,--pd_skip, or passing a boolean,--pd_skip trueor--pd_skip false.--pd_columns: Defines the loci to keep within the analysis (default when unset is to keep all loci). Formatted as a single column file with one locus name per line. For example:- Single column format
loci1 loci2 loci3
- Single column format
--pd_count_missing: Count missing alleles as different. Can be used as a flag,--pd_count_missing, or passing a boolean,--pd_count_missing trueor--pd_count_missing false. If true, will consider missing allele calls for the same locus between samples as a difference, increasing the distance counts.
The following can be used to adjust parameters for the gas mcluster tool.
--gm_thresholds: Thresholds delimited by,. Values should match units from--pd_distm(either hamming or scaled). Please see the Distance Method and Thresholds section for more information.--gm_method: The linkage method to use for clustering. Value should be one of single, average, or complete.--gm_delimiter: Delimiter desired for nomenclature code. Must be alphanumeric or one of._-.
Other parameters (defaults from nf-core) are defined in nextflow_schema.json.
To run the pipeline, please do:
nextflow run phac-nml/gasclustering -profile singularity -r main -latest --input https://github.com/phac-nml/gasclustering/raw/dev/assets/samplesheet.csv --outdir resultsWhere the samplesheet.csv is structured as specified in the Input section.
A JSON file for loading metadata into IRIDA Next is output by this pipeline. The format of this JSON file is specified in our Pipeline Standards for the IRIDA Next JSON. This JSON file is written directly within the --outdir provided to the pipeline with the name iridanext.output.json.gz (ex: [outdir]/iridanext.output.json.gz).
An example of the what the contents of the IRIDA Next JSON file looks like for this particular pipeline is as follows:
{
"files": {
"global": [
{
"path": "ArborView/clustered_data_arborview.html"
},
{
"path": "clusters/run.json"
},
{
"path": "clusters/tree.nwk"
},
{
"path": "clusters/clusters.text"
},
{
"path": "clusters/thresholds.json"
},
{
"path": "distances/run.json"
},
{
"path": "distances/results.text"
},
{
"path": "distances/ref_profile.text"
},
{
"path": "distances/query_profile.text"
},
{
"path": "distances/allele_map.json"
},
{
"path": "merged/profile.tsv"
}
],
"samples": {
}
},
"metadata": {
"samples": {
}
}
}
Within the files section of this JSON file, all of the output paths are relative to the outdir. Therefore, "path": "ArborView/clustered_data_arborview.html" refers to a file located within outdir/ArborView/clustered_data_arborview.html.
Details on the individual output files can be found in the Output documentation.
To run with the test profile, please do:
nextflow run phac-nml/gasclustering -profile docker,test -r main -latest --outdir resultsCopyright 2024 Government of Canada
Licensed under the MIT License (the "License"); you may not use this work except in compliance with the License. You may obtain a copy of the License at:
https://opensource.org/license/mit/
Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.