The simplest way to use pyCapsid is using this colab notebook which runs pyCapsid on a free Google cloud-based platform in a Jupyter environment. The Colab notebook is self-documenting and is designed to be simple to use. The Colab notebook has built in methods for visualizing the results for small structures using NGLView, but we recommend installing molecular visualization software UCSF ChimeraX locally for high-quality visualizations of larger structures.
Requires python 3.8-3.10. If you have an older version of python you may download a new version of python from here. Alternatively, you may use conda, a python package manager, to create a new virtual environment in which to install pyCapsid. You can get conda here.
If you have python 3.8-3.10 installed, simply use pip, which will install pyCapsid and all of it’s dependencies.
pip install pyCapsid
If you have a version of pyCapsid already installed, use the ‘–upgrade’ flag to update it.
pip install --upgrade pyCapsid
If you have an incompatible python version and don’t want to upgrade we recommend installing and using conda to create a virtual environment with its own python version and install pyCapsid. First, create a new virtual environment with its own Python version using conda and then activate it.
conda create -n pycapsid -y conda activate pycapsid
conda install -c luque_lab -c conda-forge pycapsid
Cupy may be used to accelerate the calculation of low-frequency modes on GPUs using CUDA. To install cupy alongside pyCapsid, use the following command to install them together:
conda install -c luque_lab -c conda-forge pycapsid cupy
This may take some time to install. For further information on installing cupy see cupy’s installation documentation. The speed improvement from GPU acceleration will depend heavily on your GPU, and will be limited by the memory available to your GPU.
For the highest quality visualization of the results we recommend downloading and installing UCSF ChimeraX.
pyCapsid has been tested on ChimeraX versions 1.5 and 1.6.
To install the necessary packages for visualizing the results in a Jupyter notebook, Via pip:
pip install jupyterlab ipywidgets==7.7.2 nglview
conda install -c conda-forge jupyterlab ipywidgets==7.7.2 nglview