gridgene.mask_properties module#

gridgene.mask_properties.timeit(func)[source]
class gridgene.mask_properties.MaskDefinition(mask, mask_name, analysis_type='per_object', grid_size=None)[source]

Bases: object

Definition of a mask to analyze.

Parameters:

  • mask (np.ndarray) – Binary mask array.

  • mask_name (str) – Name identifier for the mask.

  • analysis_type (str, optional) – Type of analysis (‘per_object’, ‘bulk’, or ‘grid’), by default “per_object”.

  • grid_size (int or None, optional) – Grid size for ‘grid’ analysis type, by default None.

mask: ndarray
mask_name: str
analysis_type: str = 'per_object'
grid_size: Optional[int] = None
class gridgene.mask_properties.MaskAnalysisResult(mask_name, analysis_type, features)[source]

Bases: object

Container for the results of mask analysis.

Parameters:

  • mask_name (str) – Name of the analyzed mask.

  • analysis_type (str) – Analysis type performed.

  • features (list of dict) – List of extracted features per object.

mask_name: str
analysis_type: str
features: List[Dict[str, Any]]
class gridgene.mask_properties.MorphologyExtractor[source]

Bases: object

Extracts morphological features from labeled masks.

extract_per_object_features(labeled_mask)[source]

Extract per-object morphological features from a labeled mask.

Parameters:

labeled_mask (np.ndarray) – Mask where each object is labeled with an integer.

Returns:

List of dictionaries containing features per object.

Return type:

list of dict

extract_bulk_features(mask)[source]

Extract bulk features for a whole mask (e.g., total area).

Parameters:

mask (np.ndarray) – Binary mask.

Returns:

Single-item list with total area and object_id=’bulk’.

Return type:

list of dict

extract_grid_features_per_object(labeled_mask, grid_size)[source]

Extract features per grid tile for each labeled object.

Return type:

List[Dict[str, Any]]

class gridgene.mask_properties.GeneCounter[source]

Bases: object

Counts gene expression values within masks.

count_genes_per_object(labeled_mask, array_counts, target_dict)[source]

Count genes per labeled object.

Parameters:

  • labeled_mask (np.ndarray) – Labeled mask array.

  • array_counts (np.ndarray) – 3D array of gene counts per pixel.

  • target_dict (dict) – Mapping from gene names to indices in array_counts.

Returns:

List of gene counts per object.

Return type:

list of dict

count_genes_bulk(mask, array_counts, target_dict)[source]

Count genes in a bulk mask.

Parameters:

  • mask (np.ndarray) – Binary mask.

  • array_counts (np.ndarray) – 3D array of gene counts.

  • target_dict (dict) – Mapping from gene names to indices.

Returns:

Single-item list with gene counts.

Return type:

list of dict

count_genes_grid_per_object(labeled_mask, array_counts, target_dict, grid_size)[source]

Count genes per grid tile within each object.

Return type:

List[Dict[str, Any]]

class gridgene.mask_properties.HierarchyMapper[source]

Bases: object

Maps child objects to parent objects based on label overlaps.

map_hierarchy(source_labels, target_labels)[source]

Map each source object ID to a list of parent object IDs from target mask.

Parameters:

  • source_labels (np.ndarray) – Labeled source mask.

  • target_labels (np.ndarray) – Labeled target mask (parent).

Returns:

Mapping from source object ID to list of parent IDs.

Return type:

dict

class gridgene.mask_properties.MaskAnalysisPipeline(mask_definitions, array_counts, target_dict, logger=None)[source]

Bases: object

Main pipeline for analyzing masks with gene counts and morphology.

run()[source]

Run the full analysis pipeline on all mask definitions.

Returns:

List of results per mask.

Return type:

list of MaskAnalysisResult

get_results_df()[source]

Get all results concatenated into a single pandas DataFrame.

Returns:

DataFrame with all extracted features.

Return type:

pandas.DataFrame

map_hierarchies(hierarchy_definitions, save_dir=None)[source]

Map child objects to their parent objects using reference labeled masks.

Parameters:

  • hierarchy_definitions (dict) – Dictionary defining the hierarchy relationships.

  • save_dir (str or None, optional) – Directory to save labeled masks, by default None.

Returns:

DataFrame with mapping of child to parent objects.

Return type:

pandas.DataFrame

This Page