spacec.archive namespace

Submodules

spacec.archive.tools_archive module

The function tl_cell_types_de performs differential enrichment analysis for various cell subsets between different neighborhoods using linear regression. It takes in several inputs such as cell type frequencies, neighborhood numbers, and patient information. The function first normalizes overall cell type frequencies and then neighborhood-specific cell type frequencies. Next, a linear regression model is fitted to find the coefficients and p-values for the group coefficient. Finally, the function returns a dataframe with the coefficients and p-values for each cell subset. The p-values can be corrected for multiple testing after the function has been executed.

spacec.archive.tools_archive.tl_Chose_window_size(windows, n_num, n_neighborhoods, sum_cols, n2_name='neigh_ofneigh')[source]
spacec.archive.tools_archive.tl_Create_neighborhoods(df, n_num, cluster_col, X, Y, regions, sum_cols=None, keep_cols=None, ks=[20])[source]
spacec.archive.tools_archive.tl_analyze_image(path, output_dir, invert=False, properties_list=['label', 'centroid', 'area', 'perimeter', 'solidity', 'coords', 'axis_minor_length', 'axis_major_length', 'orientation', 'slice'])[source]

Analyze an image by performing connected component analysis on patches and storing their information.

The function applies image processing techniques such as Gaussian smoothing, thresholding, and connected component labeling to identify and analyze patches within the image. It extracts region properties of these patches, calculates their circularity, and stores the coordinates of their contour. The resulting information is saved in a DataFrame along with a visualization plot.

Parameters:

  • path (str) – Path to the input image.

  • output_dir (str) – Directory to save the output plot.

  • invert (bool, optional) – Flag indicating whether to invert the image (default is False).

  • properties_list – (list of str): Define properties to be measured (see SciKit Image), by default “label”, “centroid”, “area”, “perimeter”, “solidity”, “coords”, “axis_minor_length”, “axis_major_length”, “orientation”, “slice”

Returns:

A tuple containing the DataFrame with region properties, including patch contour coordinates, and

the list of contour coordinates for each patch.

Return type:

tuple

spacec.archive.tools_archive.tl_apply_mask(image_path, mask_path, output_path)[source]

Apply a mask to an image and save the resulting masked image.

Parameters:

  • image_path (str) – Path to the input image.

  • mask_path (str) – Path to the mask image.

  • output_path (str) – Path to save the masked image.

Returns:

None

spacec.archive.tools_archive.tl_cell_types_de(ct_freq, all_freqs, neighborhood_num, nbs, patients, group, cells, cells1)[source]
spacec.archive.tools_archive.tl_corr_cell_ad(adata, per_categ, grouping_col, rep, sub_column, normed=True, sub_list2=None)[source]

Perform correlation analysis on a pandas DataFrame and plot correlation scatter plots.

Parameters:

  • data (pandas DataFrame) – The input DataFrame.

  • per_categ (str) – The categorical column in the DataFrame to be used.

  • grouping_col (str) – The grouping column in the DataFrame.

  • rep (str) – The replicate column in the DataFrame.

  • sub_column (str) – The subcategory column in the DataFrame.

  • normed (bool, optional) – If the percentage should be normalized. Default is True.

  • sub_list2 (list, optional) – A list of subcategories to be considered. Default is None.

Returns:

  • cmat (pandas DataFrame) – The correlation matrix DataFrame.

  • cc (pandas DataFrame) – The DataFrame after pivoting and formatting for correlation function.

spacec.archive.tools_archive.tl_generate_info_dataframe(df, voronoi_output, mask_output, filter_list=None, info_cols=['tissue', 'donor', 'unique_region', 'region', 'array'])[source]

Generate a filtered DataFrame based on specific columns and values.

Parameters:

  • df (pandas.DataFrame) – Input DataFrame.

  • voronoi_output (str) – Path to the Voronoi output directory.

  • mask_output (str) – Path to the mask output directory.

  • info_cols (list) – columns to extract from input df

  • filter_list (list, optional) – List of values to filter.

Returns:

Filtered DataFrame.

Return type:

pandas.DataFrame

spacec.archive.tools_archive.tl_generate_mask(path, output_dir, filename='mask.png', filter_size=5, threshold_value=5)[source]

Generate a mask from a maximum projection of an input image.

Parameters:

  • path (str) – Path to the input image.

  • output_dir (str) – Directory to save the generated mask and quality control plot.

  • filename (str, optional) – Name of the generated mask file (default is “mask.png”).

  • filter_size (int, optional) – Size of the filter disk used for image processing (default is 5).

  • threshold_value (int, optional) – Threshold value for binary conversion (default is 5).

Returns:

None

spacec.archive.tools_archive.tl_generate_masks_from_images(image_folder, mask_output, image_type='.tif', filter_size=5, threshold_value=10)[source]

Generate binary masks from CODEX images.

Parameters:

  • image_folder (str) – Directory that contains the images that are used to generate the masks

  • mask_output (str) – Directory to store the generated masks

  • image_type (str) – File type of image. By default “.tif”

  • filter_size (num) – Size for filter disk during mask generation

  • threshold_value (num) – Threshold value for binary mask generation

Returns:

None

spacec.archive.tools_archive.tl_generate_voronoi_plots(df, output_path, grouping_col='Community', tissue_col='tissue', region_col='unique_region', x_col='x', y_col='y')[source]

Generate Voronoi plots for unique combinations of tissue and region.

Parameters:

  • df (pandas.DataFrame) – Input DataFrame containing the data.

  • output_path (str) – Output path to save the plots.

  • grouping_col (str) – Column that contains group label that is used to color the voronoi diagrams

  • tissue_col (str) – Column that contains tissue labels

  • region_col (str) – Column that contains region labels

  • x_col (str) – Column that contains x coordinates

  • y_col (str) – Column that contains y coordinates

Returns:

None

spacec.archive.tools_archive.tl_get_distances(df, cell_list, cell_type_col)[source]
spacec.archive.tools_archive.tl_process_data(df_info, output_dir_csv)[source]

Process data based on the information provided in the DataFrame.

Parameters:

  • df_info (pandas.DataFrame) – DataFrame containing the information.

  • output_dir_csv (str) – Output directory for CSV results.

Returns:

Concatenated DataFrame of results. list: List of contours.

Return type:

pandas.DataFrame

spacec.archive.tools_archive.tl_process_files(voronoi_path, mask_path, region)[source]

Process files based on the provided paths and region.

Parameters:

  • voronoi_path (str) – Path to the Voronoi files.

  • mask_path (str) – Path to the mask files.

  • region (str) – Region identifier.

Returns:

None

spacec.archive.tools_archive.tl_spatial_context_stats(n_num, patient_ID_component1, patient_ID_component2, windows, total_per_thres=0.9, comb_per_thres=0.005, tissue_column='Block type', subset_list=['Resection'], plot_order=['Resection', 'Biopsy'], pal_tis={'Biopsy': 'orange', 'Resection': 'blue'}, subset_list_tissue1=['Resection'], subset_list_tissue2=['Biopsy'])[source]
spacec.archive.tools_archive.tl_test_clustering_resolutions(adata, clustering='leiden', n_neighbors=10, resolutions=[1])[source]

Test different resolutions for reclustering using Louvain or Leiden algorithm.

Parameters:

  • adata (AnnData) – Anndata object containing the data.

  • clustering (str, optional) – Clustering algorithm to use (default is ‘leiden’).

  • n_neighbors (int, optional) – Number of nearest neighbors (default is 10).

  • resolutions (list, optional) – List of resolutions to test (default is [1]).

Returns:

None

spacec.archive.tools_archive.tl_xycorr(df, sample_col, y_rows, x_columns, X_pix, Y_pix)[source]