Installation and Usage

Install Oris

Dependencies: Python 3.10+, boolevard, mpi4py, pandas, numpy, multiprocessing.

📝 Note: Oris is designed to run on high-performance computing (HPC) systems equipped with:

  • SLURM workload manager for job scheduling

  • An MPI implementation (e.g. OpenMPI or Intel MPI)

  • Python 3.10+ and mpi4py

To install Oris, run the following command:

pip install git+https://github.com/druglogics/oris.git

Quick Start

To execute Oris, a minimal invocation is:

oris --zips /path/to/zips --sampling 10 --mode synergies

Oris requires three mandatory parameters: --zips, --sampling, and --mode. The --zips parameter specifies the directory containint the ZIP files with the expecter internal structure (see Helpers). The --sampling parameter defiens how many models are selected from the Gitsbe-generated ensemble (see Sampling). Finally, the --mode parameter controls the operational mode of Oris, which can be set to synergies, countpaths, or full (see Oris Modes).

Oris always operates using a configuration file (see Configuration). If no configuration file is explicitly provided via the optional --config parameter, Oris automatically falls back to a default configuration.

Outputs

Oris produces outputs by augmenting the same ZIP files provided as input with an additional directory named Results. The specific contents of this directory depend on the selected operating mode (see Oris Modes):

  • synergies: writes PathCounts and SynergyExcess files .

  • countpaths: writes a PathCountsFull file.

  • full: writes PathCounts, PathCountsFull, and SynergyExcess files.