cryoridge 0.8.1

cryoRIDGE — Reliability Inferred from Density Gradient Energy for cryo-EM maps

cryoRIDGE

RIDGE = Reliability Inferred from Density Gradient Energy

Local reliability scores and build zones for cryo-EM maps from half-maps and density features. Python 3.10+.

pip install cryoridge
cryoridge          # interactive: two half-map paths → two MRC outputs
cryoridge help     # CLI reference

Interactive mode asks for half-map 1 and half-map 2, offers auto or manual contour, warns when estimated resolution is outside the model-building band (>4 Å), then writes {stem}_reliability.mrc and {stem}_build_zones.mrc under cryoridge_out/ next to half-map 1.

Pipeline (non-interactive / advanced)

Three commands, in order. Pass your own map paths and output directories.

# 1. Features from a map (.mrc / .map)
cryoridge features map.mrc --out features.npz --float32

# 2. Half-map metrics + correlations
cryoridge analyze \
  --features features.npz \
  --half1 half1.map --half2 half2.map \
  --reference ref.map \
  --contour CONTOUR \
  --out-dir analysis_out

# 3. Reliability score + build-zone MRCs
cryoridge reliability \
  --reference ref.map --half1 half1.map --half2 half2.map \
  --features features.npz \
  --contour CONTOUR \
  --out-dir reliability_out

--contour is the density threshold for the analysis mask (same value in steps 2 and 3). Step 2 (analyze) is optional if you only need the reliability and build-zone MRCs.

Reliability outputs: {label}_reliability.mrc and {label}_build_zones.mrc on your reference grid.

Flag details: cryoridge features --help, cryoridge analyze --help, cryoridge reliability --help.

HPC (ARC)

Default login python3 is too old (3.6). Conda modules load on compute nodes only:

srun -p compute1 -n 1 -t 02:00:00 --pty bash
module load miniconda/24.4.0
conda activate cryoridge    # after one-time: conda create -n cryoridge python=3.12 -y && pip install cryoridge

Batch job — copy scripts/cryoridge_cluster.sbatch.example, set MAP, HALF1, HALF2, REF, CONTOUR, and submit. Each step aborts if the previous one fails.

cp scripts/cryoridge_cluster.sbatch.example run_my_map.sbatch
# edit paths in run_my_map.sbatch
sbatch run_my_map.sbatch

If install fails with empty (from versions:), check python --version (need ≥3.10) and pip install -U pip.

Citation

@software{mohanty2026cryoridge,
  author = {Mohanty, Sarthak},
  title = {cryoRIDGE: Reliability Inferred from Density Gradient Energy},
  year = {2026},
  doi = {10.5281/zenodo.20618526},
  url = {https://doi.org/10.5281/zenodo.20618526},
  version = {0.8.1}
}

MIT — see LICENSE.