Amplicon Suite Aggregator

Description

Aggregates results from AmpliconSuite-pipeline runs.

• Accepts compressed input files (.tar.gz, .zip) of individual or grouped AmpliconSuite results, or plain directories.
• Aggregates and packages results into a new .tar.gz file, along with an aggregated .csv and .html summary.
• The output .tar.gz can be directly uploaded to AmpliconRepository.org.
• Supports batch sample renaming via a name map file.
• Can include auxiliary files alongside the upload; place them in a directory containing a file named AUX_DIR.

Installation

Option 1 — pip

pip install AmpliconSuiteAggregator

Option 2 — Git clone

git clone https://github.com/AmpliconSuite/AmpliconSuiteAggregator.git
cd AmpliconSuiteAggregator
pip install -r requirements.txt

Dependencies

Python packages: pandas, requests

Usage

python src/AmpliconSuiteAggregator.py -flist <input_list.txt> -o <project_name> [options]

input_list.txt is a plain text file with one input path per line (.tar.gz, .zip, or directory).

Aggregation options

Aggregation no longer uploads anything, and no longer runs AmpliconClassifier for you — run AC first, then aggregate its output. The -u/-s/--upload_only upload flags and the -c/--ref classifier flags were removed in 8.0.0.

Examples

Aggregate a set of results:

python src/AmpliconSuiteAggregator.py -flist input_list.txt -o MyProject

Rename samples on the way through:

python src/AmpliconSuiteAggregator.py -flist input_list.txt -o MyProject --name_map renames.tsv

Drop samples from a project you already aggregated:

python src/AmpliconSuiteAggregator.py --files MyProject.tar.gz -o MyProject_v2 \
    --exclude_samples SampleA SampleB

An aggregated archive is itself valid input, so this is how you remove a sample from an existing project: everything belonging to SampleA/SampleB is left out — their directories under results/samples/, their rows in run.json, aggregated_results.csv and every consolidated classification table, and their per-amplicon files. A name matching no sample is reported as a warning rather than failing the run.

Input files

For a breakdown of which per-sample files make up a classification project — which are essential, which are strongly recommended, and the naming patterns the Aggregator expects — see INPUT_FILES.md.

Authors

• Jens Luebeck (main contact)
• Thorin Tabor
• Edwin Huang

Issues

Bug reports and feature requests: GitHub Issues