Retention Order Support Vector Machine (ROSVM)

ROSVM is a Ranking Support Vector Machine (RankSVM) implementation for retention order prediction of liquid chromatography (LC) retention times (RT). It was initally proposed by Bach et al. (2018).

This library aims to be a more self-contained implementation, allowing the user to easily train models and make predictions.

Install

Using conda in a new environment

1. Create a new conda environment using:

   conda env create --file conda/environment.yml

2. Active the environment:

   conda activate rosvm

3. Install the package into the environment:

   pip install . 

4. (optional) Use the environment in Jupyter notebooks to run the examples:

   1. Install the IPython kernel:

      conda install ipykernel

   2. Make the environment available as notebook kernel:

      python -m ipykernel install --user --name=rosvm

Using pip

You can install the package directly using:

pip install . 

However, the installation of rdkit can be a bit tricky. You can find installation instructions for various operating systems here.

Citation

If you are using the library please cite:

• For the general approach of retention order prediction

@article{Bach2018,
    author = {Bach, Eric and Szedmak, Sandor and Brouard, Céline and Böcker, Sebastian and Rousu, Juho},
    title = "{Liquid-chromatography retention order prediction for metabolite identification}",
    journal = {Bioinformatics},
    volume = {34},
    number = {17},
    pages = {i875-i883},
    year = {2018},
    month = {09},
    issn = {1367-4803},
    doi = {10.1093/bioinformatics/bty590},
    url = {https://doi.org/10.1093/bioinformatics/bty590},
    eprint = {https://academic.oup.com/bioinformatics/article-pdf/34/17/i875/25702364/bty590.pdf},
}

• For the actual ROSVM implementation:

@software{Bach_Retention_Order_Support_2020,
    author = {Bach, Eric},
    month = {5},
    title = {{Retention Order Support Vector Machine (ROSVM)}},
    url = {https://github.com/bachi55/rosvm},
    version = {0.5.0},
    year = {2020}
}