parse

parse#

Contains functions to parse imaging data aquired on an OperaPhenix or Operetta into usable formats for downstream pipelines.

class scportrait.tools.parse.PhenixParser(experiment_dir: str | PosixPath, flatfield_exported: bool = True, use_symlinks: bool = True, compress_rows: bool = False, compress_cols: bool = False, overwrite: bool = False)#

A class to parse and manage image data from Phenix experiments.

generate_metadata() → DataFrame#: Helper function to generate metadata for the phenix experiment. Includes generating new file names for each image.

check_for_missing_files(metadata: DataFrame = None, return_values: bool = False) → list | None#

Check for missing images in the experiment. Stitching always requires a full rectangular grid of images, so any missing images (e.g. due to focus failures need to be identifed and replaced with black images).

Parameters:

metadata – DataFrame containing the metadata of the experiment. Generated by _get_phenix_metadata.
return_values – Whether to return the list of missing images or only save them in self.missing_images

Returns:

List of missing images in the experiment if return_values is True. Otherwise saves the missing images in self.missing_images.

replace_missing_images() → None#: Replace missing images with black images of the same size.

parse() → None#: Complete parsing of phenix experiment including checking for and replacing missing images.

sort_wells(sort_tiles: bool = False) → None#

Sorts parsed images according to their well.

Generates a folder tree where each well has its own folder containing all images from that well. If sort_tiles = True an additional layer will be added to the tree where all images obtained from the same FOV are sorted into a unique subfolder.

Parameters:: sort_tiles – if the images should be sorted into individual directories according to FOV in addition to well, by default False

sort_timepoints(sort_wells: bool = False) → None#

Sorts parsed images according to their timepoint.

Generates a folder tree where each timepoint has its own folder containing all images captured at that timepoint. If sort_wells = True an additional layer will be added to the tree where all images obtained from the same well are sorted into a unique subfolder according to timepoint.

Parameters:: sort_wells – if the images should be sorted into individual directories according to well in addition to timepoint, by default False

class scportrait.tools.parse.CombinedPhenixParser(experiment_dir: str, flatfield_exported: bool = True, use_symlinks: bool = True, compress_rows: bool = False, compress_cols: bool = False, overwrite: bool = False)#

Class to parse Phenix experiments where multiple experiments should be combined into one dataset. Usually this class is used if during image acquisition individual tiles were not imaged due to a focus failure. Instead of repeating the entire measurement, the missing tiles can be imaged in a separate experiment and then combined with the original dataset using this method.

This class inherits from the PhenixParser class and extends it by adding the functionality to combine multiple experiments into one dataset. These individual experiments need to be placed together in the following structure:

<experiment_name>/
└── experiments_to_combine/
    ├── experiment_1/
    ├── experiment_2/
    ├── experiment_3/
    └── ...
<experiment_name> can be chosen freely.

experiments_to_combine is a folder containing all experiments that should be combined. This folder should be placed in the main experiment directory.

experiment_n always refers to the complete folder as generated by Harmony when exporting phenix data without any further modifications.

The experiments will be combined in the order of their creation date and time. If two experiments contain images in the same position, the parser will keep the images from the first experiment.

get_datasets_to_combine() → None#: Get all phenix experiments from subdirectories that should be combined.

parse

Contents

parse#