This is a reimplementation of Pintilie et al., 2020 [1] using pure Python. It is also faster since it does not use Chimera and tries to vectorize as many things as possible. Please check the implementation to make sure the results are correct. To use this, you can run the calculate_per_residue_q_scores function in q_score.py

Install

From the repository root, create and activate a virtual environment, then install the runtime dependencies and the qscore CLI:

python3 -m venv .venv
source .venv/bin/activate
python -m pip install --upgrade pip setuptools wheel
python -m pip install -r requirements.txt
python -m pip install -e .

The editable install (-e .) makes the qscore command available while still using the source files in this checkout. For a non-editable install, use python -m pip install . instead.

Run the CLI

Check that the command is installed:

qscore --help
qscore --version

Calculate per-residue Q-scores from a structure file and cryo-EM map:

qscore calculate \
  --structure-path path/to/structure.pdb \
  --volume-path path/to/map.mrc \
  --output-path output.csv

The structure file can be PDB or mmCIF. The volume file should be an MRC file. --output-path is optional and defaults to output.csv.

1 - Pintilie, G., Zhang, K., Su, Z. et al. Measurement of atom resolvability in cryo-EM maps with Q-scores. Nat Methods 17, 328–334 (2020). https://doi.org/10.1038/s41592-020-0731-1