pyXenium.gmi.ContourGmiConfig

class ContourGmiConfig(contour_key='s1_s5_contours', contour_label_col='assigned_structure', positive_label='S1', negative_label='S5', contour_geojson=None, contour_pixel_size_um=0.2125, contour_id_key='name', feature_count=500, spatial_feature_count=100, min_cells_per_contour=20, min_library_size=1.0, min_feature_prevalence=0.05, layer=None, random_seed=1, coordinate_shuffle=False, exclude_coordinate_spatial_features=False, include_pathomics=False, inner_rim_um=20.0, outer_rim_um=30.0, rscript='Rscript', r_lib_path=None, install_gmi=True, force_reinstall_gmi=False, penalty='SCAD', lambda_min_ratio=0.02, n_lambda=100, eta=0.6, tune='EBIC', ebic_gamma=1.0, max_iter=50, spatial_cv_folds=0, bootstrap_repeats=0, bootstrap_fraction=0.8, run_label_permutation_control=False, run_coordinate_shuffle_control=False, run_spatial_feature_shuffle_control=False, run_within_label_heterogeneity=True, heterogeneity_min_contours=6, write_spatial_visualizations=True, visualization_genes=('NIBAN1', 'SORL1', 'CCND1'))

Bases: object

Configuration for canonical contour-native GMI workflows.

Parameters:

• contour_key (str)

• contour_label_col (str)

• positive_label (str)

• negative_label (str)

• contour_geojson (str | None)

• contour_pixel_size_um (float)

• contour_id_key (str)

• feature_count (int)

• spatial_feature_count (int)

• min_cells_per_contour (int)

• min_library_size (float)

• min_feature_prevalence (float)

• layer (str | None)

• random_seed (int)

• coordinate_shuffle (bool)

• exclude_coordinate_spatial_features (bool)

• include_pathomics (bool)

• inner_rim_um (float)

• outer_rim_um (float)

• rscript (str)

• r_lib_path (str | None)

• install_gmi (bool)

• force_reinstall_gmi (bool)

• penalty (str)

• lambda_min_ratio (float)

• n_lambda (int)

• eta (float)

• tune (str)

• ebic_gamma (float)

• max_iter (int)

• spatial_cv_folds (int)

• bootstrap_repeats (int)

• bootstrap_fraction (float)

• run_label_permutation_control (bool)

• run_coordinate_shuffle_control (bool)

• run_spatial_feature_shuffle_control (bool)

• run_within_label_heterogeneity (bool)

• heterogeneity_min_contours (int)

• write_spatial_visualizations (bool)

• visualization_genes (tuple[str, ...])

__init__(contour_key='s1_s5_contours', contour_label_col='assigned_structure', positive_label='S1', negative_label='S5', contour_geojson=None, contour_pixel_size_um=0.2125, contour_id_key='name', feature_count=500, spatial_feature_count=100, min_cells_per_contour=20, min_library_size=1.0, min_feature_prevalence=0.05, layer=None, random_seed=1, coordinate_shuffle=False, exclude_coordinate_spatial_features=False, include_pathomics=False, inner_rim_um=20.0, outer_rim_um=30.0, rscript='Rscript', r_lib_path=None, install_gmi=True, force_reinstall_gmi=False, penalty='SCAD', lambda_min_ratio=0.02, n_lambda=100, eta=0.6, tune='EBIC', ebic_gamma=1.0, max_iter=50, spatial_cv_folds=0, bootstrap_repeats=0, bootstrap_fraction=0.8, run_label_permutation_control=False, run_coordinate_shuffle_control=False, run_spatial_feature_shuffle_control=False, run_within_label_heterogeneity=True, heterogeneity_min_contours=6, write_spatial_visualizations=True, visualization_genes=('NIBAN1', 'SORL1', 'CCND1'))

Parameters:

• contour_key (str)

• contour_label_col (str)

• positive_label (str)

• negative_label (str)

• contour_geojson (str | None)

• contour_pixel_size_um (float)

• contour_id_key (str)

• feature_count (int)

• spatial_feature_count (int)

• min_cells_per_contour (int)

• min_library_size (float)

• min_feature_prevalence (float)

• layer (str | None)

• random_seed (int)

• coordinate_shuffle (bool)

• exclude_coordinate_spatial_features (bool)

• include_pathomics (bool)

• inner_rim_um (float)

• outer_rim_um (float)

• rscript (str)

• r_lib_path (str | None)

• install_gmi (bool)

• force_reinstall_gmi (bool)

• penalty (str)

• lambda_min_ratio (float)

• n_lambda (int)

• eta (float)

• tune (str)

• ebic_gamma (float)

• max_iter (int)

• spatial_cv_folds (int)

• bootstrap_repeats (int)

• bootstrap_fraction (float)

• run_label_permutation_control (bool)

• run_coordinate_shuffle_control (bool)

• run_spatial_feature_shuffle_control (bool)

• run_within_label_heterogeneity (bool)

• heterogeneity_min_contours (int)

• write_spatial_visualizations (bool)

• visualization_genes (tuple[str, ...])

Return type:

None

Methods

__init__([contour_key, contour_label_col, ...])
copy_with(**updates)
to_dict()

Attributes

bootstrap_fraction
bootstrap_repeats
contour_geojson
contour_id_key
contour_key
contour_label_col
contour_pixel_size_um
coordinate_shuffle
ebic_gamma
eta
exclude_coordinate_spatial_features
feature_count
force_reinstall_gmi
heterogeneity_min_contours
include_pathomics
inner_rim_um
install_gmi
lambda_min_ratio
layer
max_iter
min_cells_per_contour
min_feature_prevalence
min_library_size
n_lambda
negative_label
outer_rim_um
penalty
positive_label
r_lib_path
random_seed
rscript
run_coordinate_shuffle_control
run_label_permutation_control
run_spatial_feature_shuffle_control
run_within_label_heterogeneity
spatial_cv_folds
spatial_feature_count
tune
visualization_genes
write_spatial_visualizations

contour_key: str = 's1_s5_contours'

contour_label_col: str = 'assigned_structure'

positive_label: str = 'S1'

negative_label: str = 'S5'

contour_geojson: str | None = None

contour_pixel_size_um: float = 0.2125

contour_id_key: str = 'name'

feature_count: int = 500

spatial_feature_count: int = 100

min_cells_per_contour: int = 20

min_library_size: float = 1.0

min_feature_prevalence: float = 0.05

layer: str | None = None

random_seed: int = 1

coordinate_shuffle: bool = False

exclude_coordinate_spatial_features: bool = False

include_pathomics: bool = False

inner_rim_um: float = 20.0

outer_rim_um: float = 30.0

rscript: str = 'Rscript'

r_lib_path: str | None = None

install_gmi: bool = True

force_reinstall_gmi: bool = False

penalty: str = 'SCAD'

lambda_min_ratio: float = 0.02

n_lambda: int = 100

eta: float = 0.6

tune: str = 'EBIC'

ebic_gamma: float = 1.0

max_iter: int = 50

spatial_cv_folds: int = 0

bootstrap_repeats: int = 0

bootstrap_fraction: float = 0.8

run_label_permutation_control: bool = False

run_coordinate_shuffle_control: bool = False

run_spatial_feature_shuffle_control: bool = False

run_within_label_heterogeneity: bool = True

heterogeneity_min_contours: int = 6

write_spatial_visualizations: bool = True

visualization_genes: tuple[str, ...] = ('NIBAN1', 'SORL1', 'CCND1')

to_dict()

Return type:

dict[str, Any]

copy_with(**updates)

Parameters:

updates (Any)

Return type:

ContourGmiConfig