TOGA2

TOGA2: A faster, more versatile successor of Tool to infer Orthologs from Genome Alignments

Important

TOGA2 is currently in early access phase. This means, certain TOGA2 features and most of the documentation at TOGA2 wiki are currently missing and will be added in the following days. If you want to use the early access version and experience problems with installing or running code in this repository, please open an issue here or contact the TOGA2 team (yury.malovichko@senckenberg.de).

Documentation

Detailed explanations of all output files can be found in our TOGA2 Wiki.

Changelog

v2.0.7

• run mode:
  • Replacing positional arguments with keyword arguments
  • --isoform_file, --u12_file, and --spliceai_dir options are now "semi-mandatory"; the user is expected to provide the respective arguments unless the explicit deprecative flags are set
  • Alternative input formatting with --input_directory, --ref_name, and --query_name shortcuts: Format your data storage tree once and enjoy simplified command line interface
  • Postoga summary table (toga.table.gz) added to the output for run mode
  • Projections of the same reference gene/transcript overlapping by absolute coding sequence coordinates are now collapsed into a single query gene regardless of their overlap by coding exon coordinates
• NEW MODE: postoga for Postoga integration
• NEW MODE: sequence-alignment for orthologous sequence alignment across multiple same-referenced TOGA2 runs (alpha version)
• Apptainer support (see supply/containers):
  • Stable local execution container image
  • Batch manager-compatible image template
  • Removing toga2.py as a container entry point
  • Adding container support for parallel step scripts (see supply/containers/README.md)
• Updated local installation
  • Postoga installation
  • Conda environment support
  • Updated bigWigToWig version (-bed and -header options) now distributed with TOGA2

For the full list of code changes, see changelog.md## Installation

from Github

TOGA2 Makefile provided in this repository contains directives for code compilation, third party software installation, Python package download, and model training.

git clone --recurse-submodules https://github.com/hillerlab/TOGA2
cd TOGA2
make

Important

Executing the Makefile will not install Nextflow and Cargo on your machine.

Since the Make directive also installs Python packages globally, you might want to create a dedicated virtual environment beforehand (Python version 3.13 or higher):

git clone --recurse-submodules https://github.com/hillerlab/TOGA2
cd TOGA2
python3 -m venv toga2
source toga2/bin/activate
make

As an alternative to Python virtual environment, you can also use Conda for environment creation. TOGA2 comes with a Conda configuration file which creates an environment with Python, Cargo, and Nextflow:

git clone --recurse-submodules https://github.com/hillerlab/TOGA2
cd TOGA2
conda env create -f conda.yaml
conda activate toga2
make

via Apptainer

Container image definition file for Apptainer is provided with TOGA2 under supply/containers/apptainer.def. To build the container, make sure you have Apptainer installed, then run the following commands:

git clone --recurse-submodules https://github.com/hillerlab/TOGA2
TMPDIR=${tmp_dir} apptainer build ${container.sif} supply/apptainer.def

where:

• ${tmp_dir} is the path to the directory to store Nextflow temporary files in;
• ${container.sif} is the path to save the resulting image to.

The resulting container has toga2.py as an entry point. If you run the following or similar command:

TMPDIR=${tmp_dir} apptainer run --bind ${bound_dir1},${bound_dir2} ${container.sif} toga2.py

you should see the TOGA2 start menu.

Note

The image provided in supply/ directory contains the latest TOGA2 release, third-party software used for input preparation and TOGA2 annotation, and Nextflow for parallel process management. The container, however, does not contain any Nextflow-compatible parallel job executor. To set up your container for batch manager compatibility, see README at supply/containers and example recipe at supply/containters/apptainer_slurm.def

Running TOGA2

If activated without additional arguments (toga2.py), the following start screen is displayed in the user's terminal:

Usage: toga2.py [OPTIONS] COMMAND [ARGS]...

  MMP""MM""YMM   .g8""8q.     .g8"""bgd      db          `7MMF'`7MMF'
  P'   MM   `7 .dP'    `YM. .dP'     `M     ;MM:           MM    MM  
       MM     dM'      `MM dM'       `     ,V^MM.          MM    MM  
       MM     MM        MM MM             ,M  `MM          MM    MM  
       MM     MM.      ,MP MM.    `7MMF'  AbmmmqMA         MM    MM  
       MM     `Mb.    ,dP' `Mb.     MM   A'     VML        MM    MM  
     .JMML.     `"bmmd"'     `"bmmmdPY .AMA.   .AMMA.    .JMML..JMML.

  TOGA2 - Tool for Ortholog Inference from Genome Alignment

Options:
  -V, --version      Show the version and exit.
  -help, -h, --help  Show this message and exit.

Commands:
  cookbook            List example commands for 'run' mode
  from-config         Run TOGA2 pipeline with a configuration file
  integrate           Prepare an integrated TOGA2 annotation  by combining annotation with different references
  merge               Merge complementing TOGA2 results for the same reference and query
  postoga             Run postprocessing analysis with Postoga
  prepare-input       Prepare reference annotation files for TOGA2 input
  run                 Run TOGA2 pipeline with command line arguments
  sequence-alignment  Align orthologous sequences from multiple TOGA2 results
  spliceai            Generate SpliceAI predictions for query assembly
  test                Test TOGA2 with companion dataset

Except for toga2.py test, invoking any of the listed commands without arguments also displays the help message. You can also invoke help message for TOGA2 or any of its commands with --help/-h option.

Test run

To ensure that TOGA2 was installed properly, run the following command:

toga2.py test

Provided the successful installation, you will see the TOGA2 execution log printed to stdout, with the results stored at TOGA2/sample_output.

Note

If you are running TOGA2 from Apptainer, you might have to provide a custom output directory with --output/-o option to bypass the read-only container configuration:

TMPDIR=${tmp_dir} apptainer run --bind ${bound_dir1},${bound_dir2} ${container.sif} toga2.py test -o ${output_dir}