--resegment

Reprocess G4X output with a new segmentation

Replaces or updates the segmentation mask in a G4X run and regenerates all downstream single-cell data and .bin files.

usage

---

$ resegment
  --run_base /path/to/G4X/output 
  --segmentation_mask /path/to/new_mask.npz 

  # ─── optional ───
  --sample_id <sample_id> 
  --out_dir <output_dir> 
  --segmentation_mask_key <mask_array_name> 
  --threads <n_threads> 
  --verbose <level>

argument descriptions

---

required

--run_base: (type: str)

Path to the G4X sample output folder (the base directory for the run). This directory must contain required files such as run_meta.json, segmentation masks, and panel files.

--segmentation_mask: (type: str)

Path to the new segmentation mask file. Supported formats include .npy, .npz, and .geojson. This file will be used to replace the existing mask for transcript and protein signal assignment.

optional

--sample_id: (type: str default: None)

Optional sample identifier. If not provided, the sample ID will be inferred from the name of the run_base directory.

--out_dir: (type: str default: None)

Directory to write the updated segmentation and downstream output files. If not provided, existing files in the run_base directory will be overwritten in-place.

--segmentation_mask_key: (type: str default: None)

Specifies the identifier for segmentation labels when loading mask data:

• If using a .npz file:
  Provide the name of the array within the archive that corresponds to the segmentation mask (required if multiple arrays are stored).

• If using a .geojson file:
  By default, cell labels are expected in a column named label.
  Use this argument to override and select a different column as the label source.

--threads: (type: int default: 4)

Number of threads to use for segmentation, signal extraction, and downstream computation.

--verbose: (type: int default: 1)

Logging verbosity level.
This affects how much information is printed to the console during execution.

• 0 = WARNING
• 1 = INFO
• 2 = DEBUG