cryoem-halfmap-qc 0.3.3

Half-map QC and local reliability scores for cryo-EM density maps

cryoem-halfmap-qc

Half-map reproducibility and local reliability scores for cryo-EM density maps: density features, windowed half-map correlation, reliability_score, and build / caution / omit zones.

Install the halfmap-qc command:

pip install cryoem-halfmap-qc
halfmap-qc --version

From source:

pip install "git+https://github.com/sarthaktexas/cryoem-halfmap-qc.git@v0.3.3"
# or: git clone … && pip install -e .

---

Quick start

Run from a working directory with your maps (and optionally cohort/manifest.csv for batch mode).

Interactive menu (terminal):

halfmap-qc
halfmap-qc help

Single map — features:

halfmap-qc features map.mrc --out features.npz --float32

Half-map analysis + reliability export:

halfmap-qc analyze \
  --features features.npz \
  --half1 half1.map --half2 half2.map \
  --reference deposited.map --contour 0.116 \
  --out-dir outputs/analysis

halfmap-qc reliability --emd-id 49450 --contour 0.116 \
  --features features.npz \
  --halfmap-npz outputs/analysis/halfmap_metrics.npz

Cohort batch (uses cohort/manifest.csv + local data/ paths):

halfmap-qc cohort --pending
halfmap-qc cohort --emd-id 49450
halfmap-qc cohort-ids    # EMDB IDs for SLURM array jobs

---

CLI

Command	Purpose
halfmap-qc	Interactive menu (TTY)
halfmap-qc help	Command reference
halfmap-qc features	Local density / multiscale features → .npz
halfmap-qc analyze	Windowed half-map CC + feature correlations
halfmap-qc reliability	Reliability score, build zones, MRC export
halfmap-qc cohort	Batch pipeline from manifest
halfmap-qc cohort-ids	List cohort EMDB IDs

Per-command flags: halfmap-qc cohort --help, etc.

---

Data layout

Maps are not stored in this repository. Typical layout:

data/emd_<ID>-<label>/     # deposited map + half-maps
outputs/emd_<ID>/           # pipeline outputs (created by halfmap-qc)
cohort/manifest.csv         # optional batch manifest

Download maps from EMDB. Use each entry's depositor-recommended contour in the manifest.

---

Python API

from cryoem_mrc import load_full_and_half_maps, run_pipeline, half_map_local_metrics
from cryoem_mrc.reliability import compute_reliability_maps, classify_build_zones

bundle = load_full_and_half_maps(
    "deposited.map", "half1.map", "half2.map",
    dtype="float32", resample_if_needed=True,
)
metrics = half_map_local_metrics(bundle.half1.data, bundle.half2.data, window=5)

features = run_pipeline("avg.map", use_float32=True)
delta = bundle.half1.data - bundle.half2.data
rel = compute_reliability_maps(
    features["density_normalized"], delta, window=5, mask=contour_mask
)
zones = classify_build_zones(rel["reliability_score"], contour_mask)

Main modules: io, map_grid, local_stats, half_map_repro, reliability, analysis.

---

What the scores mean

• Windowed half-map correlation — local Pearson correlation between half-maps in a sliding window (fast reproducibility proxy).
• reliability_score — in-mask percentile rank of the constraint term V (gradient-based smoothness vs half-map disagreement); higher = more reliable for modeling.
• Build zones — tercile labels inside the density contour: omit / caution / build.

External local-resolution maps (BlocRes, ResMap) are optional comparison inputs, not required to run the pipeline.

---

Tests

python -m unittest discover -s tests -q

---

Citation

@software{mohanty2026cryoem_halfmap_qc,
  author = {Mohanty, Sarthak},
  title = {cryoem-halfmap-qc: local map reliability from cryo-EM density and half-maps},
  year = {2026},
  doi = {10.5281/zenodo.20618526},
  url = {https://doi.org/10.5281/zenodo.20618526},
  version = {0.3.3}
}

See also CITATION.cff for GitHub's Cite this repository button.

---

License

MIT — see LICENSE.