changelog.txt

2.17.1
  - Major: New docker image: Updated gtdbtk to v2.5. (Still using GTDB release 226)
  - Using europe based mirror for downloading the gtdb data package. 
  - Minor changes to gather.smk to avoid errors when partial downloads exist. 
  - Added new rule: bootstrap_mashtree which uses Katz' internal perl script. 

2.16.2
  - Fixed wrong database path for bakta and antismash

2.16.1
  - 2nd try: Fixed some versions in the conda environments (has little to no effect on the overall pipeline). All conda environments now solve in strict channel priority mode.
  - Database downloads into a version-minor named directory inside environment variable $COMPAREM2_DATABASES
  - Updated checkm2 database and pinned the conda env to the minimum version that supports this database.
  - Updated eggnog to v2.1.13

2.15.3
  - Reverted refactored workflow envs (.yaml) as the previous release (v2.15.2) was a patch and not a minor release.

2.15.2
  - New add_ncbi batch downloader with advanced caching.
  - Fixed snakemake keep-incomplete to false.
  - Fixed some versions in the conda environments (has little to no effect on the overall pipeline). All conda environments now solve in strict channel priority mode.
  - Updated CI tests so they use strict channel priority.
  - Added escaped parentheses to the  --gram "?" setting in bakta as it otherwise triggers an error when running with docker/apptainer.


2.15.1
  - Breaking change: Adding new conda environment to docker image containing the ncbi datasets cli tool.
  - Add option to add ncbi analysis as comparison to local genomes using `comparem2 --config add_ncbi=<accession>`
  - Changed the way --status works, so now it computes the summary ahead of time. #122
  - Fixed an error where some sample basis jobs would rerun when other samples were added or removed.
  - Added more in-depth documentation on the --until parameter
  - Set Bakta as the default annotator (previously prokka, which is now deprecated)
  - Report graphics are now exported as pdf (vector) into the results directory.
  - Fixed bug regarding the snp-dists section in the report.
  

2.14.1
  - Breaking change: New Docker image.
  - Added gapseq, rule "gapseq_find" is done and ready.
  - Updated gtdbtk to release 226.
  - Minor fixes.

2.13.1
  - Breaking change: New Docker image.
  - Added amrfinder, removed abricate.
  - Added text hints.
  - Add version info to report.
  - New checkpoint system that uses file testing and failures instead of the built in checkpoint system which turned out to not be working correctly. @AstaLaugesen
  - Deeper CI tests on GitHub Actions.
  

2.12.1
  - Breaking change: New Docker image.
  - Remove anaconda and defaults channels from worklow conda envs.
  - Rule "all" inputs are now set by the number of input genomes.
  - Added --downloads functionality. (Thanks to @avera1988)
  - Handles empty input genomes.
  - Removed busco altogether. Had too many problems with hanging jobs and crashes/incomplete results. 
  

2.11.2
  - Disabled any2fasta. Not relevant when only supporting fasta files is input anyway.
  - Updated KEGG data from 2023-08-28 to 2024-08-13.
  - Added --status functionality.
  - Minor streamlining of code.
  - Hard-coded eggnog decorate gff "yes" (#114)
  - Made the panaroo rule and report section robust to a missing alignment file when the core is empty. 

2.11.1
  - Speed up: Eggnog reuses the proteins called by the default annotation tool instead of redoing it itself with prodigal. 
  - Minor fixes.
  - Enable antismash (still only works when using Conda and not Apptainer)
  - Docker image did not contain bakta when annotator was set to prokka during containerization. Issue fixed within v2.10.1. Thanks to @flefler.
  - Passthrough parameters notice removed.

2.10.1
  - Breaking change: New Docker image (pinned eggnog)
  - Variation in output files from Eggnog means that eggnog is now pinned to v2.1.12. Thanks to @michoug
  - Quiet report output
  - Minor fixes.

2.9.1
  - Breaking change: new passthrough parameter system rolled out to most rules.
  - Update cite value to reflect preprint doi https://doi.org/10.1101/2024.07.12.603264 
  - Add more tools to passthrough argument feature. And write documentation.
  - Fixed an error that meant that panaroo was using the Prokka annotation (.gff) regardless of the annotator being set with --config annotator=[bakta|prokka].
  - Fixed an error that meant that both Bakta and Prokka would run even though the user had selected one with --config annotator=[bakta|prokka]. Thanks to @michoug
  - Minor fixes.

2.8.2
  - Add --cite parameter to executable.
  - New more differentiated logo for cli.
  - Prints report sections in the end of the dynamic report pipeline.
  - Explicit option to run prokka in archaea mode.
  - config setting prokka_rfam is false by default.
  - Brought Panaroo options out into config.
  - Docker image follows minor version only. So all subsequent patches will use the same Docker image for Apptainer.
  - New passthrough parameter system where parameters and arguments are automatically unpacked for each rule. Generalized implementation.
  - Minor fixes.

2.8.1
  - Breaking change: Changed named from Assemblycomparator2 to CompareM2.
  - Compressed test files.
  - Updated header text.
  - Bring back the ass.
  - Add --no-check-certificate to wget downloads.
  - Added missing localrules for download rules.
  - In profiles, decrease latency and add rerun-incomplete: true.
  - Increase memory requirement for Bakta.
  - Use dbcan_build to install that database.
  - Disable antismash until issue is resolved.
  - Updated the busco report section so it accepts the .faa file that was added in v2.7.1.
  - Add tmpdir to eggnog.
  - Compute md5sum of input genome in rule copy.
  - Minor fixes.

2.7.1
  - Breaking change: File hierarchy now follows what is recommended in the snakemake documentation. Rules are now in individual files.
  - Breaking change: profiles/ moved to profile/ (sorry). And local profiles are now named "default".
  - Busco runs in protein mode, reusing .faa from prokka (or bakta - see further down). Should result in speedup.
  - Added Panaroo, removed Roary.
  - For dbcan, only the substrates are shown in the dynamic report.
  - Gtdb updated to release 220.0
  - Add bakta.
  - User can specify whether to use prokka or bakta to create annotation that is used for dbcan, interproscan diamond_kegg, busco. This can be set using `--config annotator=bakta` or by changing the permanent parameter in config/config.yaml.
  - Add eggnog
  - rename benchmark files from individual->Sample
  - Add antismash. Uses .gbk from prokka.
  - report_subpipeline renamed to dynamic_report
  - Dynamic report is now run by the ./comparem2 binary instead of using onsuccess/onerror.
  - Minor fixes.
  - Software versions and database versions are written to hidden files within output directories.
  - Added dynamic parameters to prokka and iqtree to the config file.
  - Added support for a "fofn" which is a file of file names. In the tseemann style, it defines which input genomes to process.
  - Use KEGG KO from eggnog in kegg_pathway. If this doesn't work well I'll consider re-implementing kofam_scan which was removed long time back. Now a_class and b_class are "class" and "group".
  - Implemented a way to set the output directory which now defaults to the old static "results_comparem2".
  - Implemented TreeCluster that on the 4.5% and 10% basis defines clusters on the mashtree distances so the user can easily get an impression on the relative differences between samples.
  - Fix an error that makes snp-dists results not show up in the report. (Thanks to @AstaLaugesen)
  - Added documentation over at https://comparem2.readthedocs.io/en/latest/

2.6.2
  - Fix wrong path for dbcan results (not yet implemented in report)
  - Downgrade diamond 2.1.8 due to segfault in newest version (2.1.9) https://github.com/bbuchfink/diamond/issues/785

2.6.1
  - Remove Kraken2, hence new minor version. Kraken2 really only makes sense on reads. 
  - Roary as a checkpoint, to only continue with iqtree, fasttree and snp-dists if the core genome is non-empty.
  - conda yaml for wget. TODO, carry over to dbcan download as well.

2.5.18
  - Support for weird characters in input files. #31
  - Support for adding input files from other directories. #58
  - Minor bugfixes. #77
  - Test battery is being developed. (https://github.com/cmkobel/comparem2-test-battery)

2.5.17
  - Forced recompilation of docker image did not fix issue #74. This patch has equivalent .yaml-files to version 2.5.14, which means that the old (working) docker image will be compatible.

2.5.16
  - Indirect bug because of an uncaught docker error. Forced recompilation of the image should fix that.
  - Propagation of version string to easy version bumping.

2.5.15
  - Limited snakemake version because of breaking dependency changes herein.

2.5.14
  - problems with dockerfile fixed

2.5.13
  - Bugfix: Wrong number of bootstraps for quicktree

2.5.12
  - motulizer.yaml was missing.

2.5.11
  - Added support for motulizer and motupan. Yet lacking in report and awaits further integration.

2.5.10
  - comparem2 now supports iqtree with bootstrapping. Runs on the core genome produced by roary. Yet to be included in the report.

2.5.9
  - Fixed an error where checkm2 would fail setting the download location internally when running comparem2 using the apptainer/docker container. https://github.com/chklovski/CheckM2/issues/73

2.5.8
  - Fixed an error where a different version of the docker image would be pulled for the report subpipeline
  
2.5.7 
  - Fixed an error where a conda environment would not be activated when using the apptainer profile for slurm.

2.5.6
  - New ./comparem2 "binary" to replace the old alias. This "binary" comes with default environment variables.

2.5.5
  - Use $ASSCOM_DATABASES to define where the databases are placed. 
  - Docker images have a versioning system so an incompatible image can't be used.

2.5.4
  - Using the mashtree order in plots in the report. Using alphabetical order redundantly in case mashtree hasn't run (sucessfully). Minor adjustments to other plots. TODO: midpoint root tree somehow and order all tables in the report as well.

2.5.3
  - Using snakemake's --containerize to automatically generate images which are then synced to dockerhub as cmkobel/comparem2. Support for hpc queues isn't yet tested.

2.5.2
  - dbcan, including download
  - interproscan
  - both dbcan and interproscan are yet to be integrated in the report

2.5.0
  - Apptainer support. New profiles for apptainer or conda use.

2.4.1
  - Adjusting the new report and fix minor bugs

2.4.0
  - New report based on child sections

2.3.0
- Disabled pathway enrichment as conda failed
- Implemented GTDB-tk
- Having problems with Prokka

2.2.0
- Sample pathway enrichment analysis with clusterProfiler which is integrated into the report.
- Considering going away from singularity: Too much maintenance and too little return when it has to work with conda anyway. I haven't completely decided yet, but leaning strongly.

2.1.1
- Simple GC statistics added to output and report 

2.1.0
- Bug and stability issues fixed

2.0.0
- Completely new system based on snakemake using both slurm and singularity