- Added antimicrobial resistance genes with AMRplusplus.
- An experimental function to highlight the abundance of taxa of interest. The first experimental organism list is wastewater related organisms obtained from LA County Sanitation District.
- Sourmash output now uses sequence IDs (assembly accession numbers) as keys. Pipeline now generates a Qiime2 feature table with these keys, in addition to the table with collapsed taxonomy.
- MultiQC report has a title that matches the run name. User can use
-name
to change the report title.
- Visualizations for taxonomy profiling results in the report.
- Diversity analysis using Qiime2. Including exporting taxonomy results to Qiime2, merge/filter samples, alpha rarefaction, alpha/beta diversity.
- Diveristy visualizations including heatmaps, alpha rarefaction curve, alpha/beta diversity plots.
- Filtering of samples or taxa based on low read count, minimum reads per taxa, minimum sample per taxa.
- New profiler: sourmash.
- Zymo's sourmash database, which contains host sequences. Added support scripts to separate host reads from rest, and correspoding MultiQC visualizations.
- Optional khmer trim-low-abund.py for sourmash.
- Compare with a reference dataset using Qiime2.
- A process to generate an output manifest JSON file, part of compatibility requirements for Aladdin.
- Metaphan 3 changed to metaphlan 4
- User must choose one profiler-database combination, instead of running multiple profilers. Temporarily disabled profilers except sourmash and metaphlan4.
- Changed the format of sample sheet, aka. design CSV file and the code that sanity checks this file.
- Temporarily disabled support for Nanapore data.
- Tie read length filtering in preprocessing to kmer size, when sourmash is selected.
- Ignore samples that failed read QC or profiling. Let pipeline finish successfully regardless of a few bad samples. Add warnings for failed samples to report.
- Hide most parameters from Aladdin users.
- Use Aladdin report template.
- Eliminated 'perform_shortread_qc' and 'perform_shortread_complexityfilter' paramters and merged them into others.
- Sample name cleaning in report. Now one sample occupy one row in the general stats table.
- FastQC results not correctly fed to MultiQC for single-end data.
- nf-core related files and code that are not required any more.
The format is based on Keep a Changelog and this project adheres to Semantic Versioning.