What is this?

This is the standalone version of OutCyte and here is the paper.

Dependencies

• Python 3.7
• Core package dependency: PyTorch 1.0.1 (can be replaced with Numpy implementations), XGBoost 0.9
• We consider you are using a terminal and know basic Python and bash commands. For windows users, you may need to install a linux-based terminal, such as MobaXterm.

How to

1. Clone the repo or download it and unzip the software to a folder named "outcyte"

2. Change working directory to "outcyte".

3. Set up a virtual environment with conda : conda create -n outcyte --file requirements.txt, then activate the python virtual environment by running the following command: conda activate outcyte. Remember to conda deactivate when you finish.

4. Now you can annotate your fasta file by typing:

python run_outcyte.py /path/to/your/fasta/file <outcyte-model> <output_dir>

where "outcyte-model" specifies the method you would want to use, so it could be either one of "outcyte-sp", "outcyte-ups" or "outcyte". It is basically the same as the web version.

5. The results will be stored in the subfolder you define as < output_dir >.