Skip to content

Install

Jump to bottom
Wout Bittremieux edited this page Mar 28, 2022 · 5 revisions

Requirements

  • Python version: ANN-SoLo requires Python version 3.6 to 3.9. Python 3.10 and newer are currently not supported yet.
  • Operating system
    • CPU version: Linux and OS X.
    • GPU version: Linux, using an NVIDIA CUDA-enabled GPU device.

Installation

We recommend using conda to create a separate environment for ANN-SoLo:

conda create -n ann_solo python=3.7

Next, activate this environment:

conda activate ann_solo

NumPy needs to be available before ANN-SoLo can be installed, whereas the other dependencies can be automatically installed while you install ANN-SoLo. Here we will explicitly install all dependencies from the appropriate conda channels simultaneously:

conda install configargparse cython joblib matplotlib numba numexpr numpy pandas pyteomics scipy spectrum_utils tqdm -c defaults -c bioconda -c conda-forge

The mmh3 library is not available as a conda package but can be installed using pip:

pip install mmh3

The Faiss installation depends on a specific GPU version. Please refer to the Faiss installation instructions for more information.

To install the CPU-version of Faiss:

conda install faiss-cpu -c pytorch

To install the GPU-version of Faiss (please make sure your GPU supports the appropriate CUDA version):

conda install faiss-gpu cudatoolkit=10.0 -c pytorch

Now we can install ANN-SoLo using pip:

pip install ann_solo

You can verify whether ANN-SoLo has been installed correctly using the ann_solo command:

ann_solo -h

Congratulations! You can now start using ANN-SoLo to process your proteomics data.

Dependencies

ANN-SoLo has the following dependencies:

See above for the recommended way to install these dependencies using conda. Any missing dependencies will be automatically installed when you install ANN-SoLo via pip.

Clone this wiki locally