v3.0.0

What's Changed

• Revise workflow input fields and metadata I/O by @etal in #50
• Restructure the report and output TSV schema by @etal in #53
• CI automation to build and deploy the Docker image by @etal in #60, #61, #62, #63

Full Changelog: v2.6...v3.0.0

v2.6

What's Changed

• Expose more options in workflow; clean up core Python scripts by @etal in #10
• Update in-repo documentation by @etal in #13
• test: Ensure build/ output directory exists before writing to it by @etal in #15
• Refactor report template vs. scripts by @etal in #14
• Report updates and bugfix by @etal in #27
• Add FASTQ input and position-sorted, indexed BAM output by @etal in #28
• Use a second sentinel file in case 2 input files are missing by @etal in #33

Full Changelog: v2.5...v2.6

v2.5

Version 2.5

This release includes a standard Nextflow workflow that takes unaligned PacBio HiFi
reads from AAV sequencing as input, aligns them to genome reference sequences in the
recommended way, and then performs processing as before to generate the AAV report.

This workflow includes configuration to run locally, on Google Cloud, and on the Form
Bio platform. The individual Nextflow processes (under modules/local/laava.nf) are
designed to be easily included in other Nextflow workflows, for extensibility and
customization.

Packaging

• Dockerfiles are provided to create images 'laava', for typical use within the
  workflows and interactively, and 'laava_dev', for development. The latter excludes the
  AAV processing scripts and includes more dependencies. The corresponding conda
  environments have matching names.
• The 'laava' container image is now publicly hosted on GitHub Packages, linked to this
  source repo.
• The semi-automated test suite includes a small BAM file in the repo, downsampled from
  the public PacBio scAAV example dataset. This enables end-to-end testing of the
  pipeline.

Processing scripts

• New script prepare_annotation.py generates the specialized "annotation.txt" file
  from vector annotations in standard BED format and a simple list of additional
  non-vector sequence names (e.g. host genome, helper and repcap plasmids).

Report

• Updated the overview figures to include a "snapback" classification for certain reads
  in ssAAV samples.
• Updated the text in the AAV type/subtype definition tables, also including the
  "snapback" read-level definition.
• In the first two data tables, list "Frequency in AAV" and "Total Frequency" as
  separate columns.
• Resolved a few quirks in rendering tables and conditional subsections.
• Deleted legacy report generation script plotAAVreport.R.
• Updated the "Methods" section text to describe the end-to-end workflow.

Full Changelog: v2.1...v2.5

v2.1

This release revamps the output report and improves several aspects of portability. Script usage is similar to previous releases.

Report updates

• A new script create_report.R, intended to replace plotAAVreport.R, generates equivalent reports in both HTML and PDF format from the same Rmarkdown source. Command-line usage of create_report.R is the same as plotAAVreport.R, and it should function as a drop-in replacement. (#3)
• The updated report includes new plots, tables, and explanatory text regarding AAV types, subtypes, and flip/flop configurations. (#3, #5)

Portability

• Docker: the scripts and dependencies are captured in a Dockerfile that can be used to build portable containers. (#1)
• Automated testing: a simple CI suite now runs on each PR via GitHub Actions to verify the code under Python 3.8, 3.9, and 3.10. (#1)
• MIT license: the codebase is now formally available as open-source software under a permissive, OSI-compatible MIT license. Previously, the permissible use was not stated explicitly. (#2)
• Semi-manual testing: a new test/Makefile can be run manually to pull in public PacBio example datasets and run this repo's scripts on them to perform the analysis and produce reports. (#3)