bfastor 0.1.0

Fast cryo-EM B-factor refinement

B-FASTor

Rapid refinement of atomic B-factors for cryo-EM derived models using the Jax library.

Installation

To install the CPU only version of B-FASTor run:

pip install bfastor

To install B-FASTor configured to run on NVIDIA GPUs, run:

pip install bfastor[cuda12]

If you have access to Google's TPU accelerators, it is possible to install a TPU compatible version of B-FASTor with the commands:

pip install -U "jax[tpu]" -f https://storage.googleapis.com/jax-releases/libtpu_releases.html
pip install bfastor

Usage

B-FASTor can be used a CLI application, via the bfastor-refine command. This can be run on any map/model combination using the following command:

bfastor-refine --model path/to/file.pdb --map path/to/map.mrc

The output files, including plotting B-factors, and model with refined B-factors will be saved in a folder with the name of the input model file, e.g. for the model 7kx7.pdb the output will be saved in 7kx7/.

Support

If you have any issues please raise them in the issues page.

Authors and acknowledgment

B-FASTor is developed and maintained by Joseph Beton, working within Professor Maya Topf's research group at the CSSB/Leibniz Institute for virology in Hamburg, Germany.

License

B-fastor is licensed under the MIT-license. The full license is here