stitching¶

Find or apply coordinates to stitch an image

pysmFISH.stitching_package.stitching.assess_performance(micData, alignment, joining, cov_signal, xcov_list, folder, use_IJ_corners=False)[source]¶

Assess the performance of the stitching

This functions writes its the result to a file in “folder”.

micData: object: MicroscopeData object. Contains coordinates of the tile corners as taken from the microscope and contains the tile set.
alignment: dict: Contains key ‘P’ with a 1D numpy array containing pairwise alignment y and x coordinates (and z-coordinates when applicable) for each neighbouring pair of tiles, array will be 2 * len(contig_typles) for 2D data or 3 * len(contig_typles) for 3D data. Also contains key ‘covs’ with a 1D numpy array containing covariance for each pairwise alignment in ‘P’, ‘covs’ will be len(contig_typles).
joining: dict: Contains keys corner_list and final_image_shape. Corner_list is a list of list, each list is a pair of a tile index (int) and it’s tile’s shifted coordinates in the final image (numpy array containing floats). Final_image_shape is a tuple of size 2 or 3 depending on the numer of dimensions and contains ints.
cov_signal: np.array: The covariance of each neighbouring tile pair in the part of the tiles that overlap in the final stitched signal image.
xcov_list: list: List of cross covariance of the overlap of the tiles in the final stitched image. As returned by tilejoining.assess_overlap.
folder: str: String representing a path. The folder where the performance report should be saved. Needs a trailing slash (‘/’).
use_IJ_corners: bool: If True compare our corners to Image J found in a file in folder, the file name should contain: TileConfiguration.

pysmFISH.stitching_package.stitching.get_pairwise_alignments(tiles, tile_file, contig_tuples, micData, nr_peaks=8, nr_slices=None, nr_dim=2)[source]¶

Calculate the pairwise transition

Calculates pairwise transition for each neighbouring pair of tiles. This functions is only used in the single core version of the code, not when using MPI.

tiles: list: List of references to the the tiles in the hdf5 file tile_file.
tile_file: pointer: HDF5 file handle. Reference to the opened file containing the tiles.
contig_tuples: list: List of tuples. Each tuple is a tile pair. Tuples contain two tile indexes denoting these tiles are contingent to each other.
micData: object: MicroscopeData object. Containing coordinates of the tile corners as taken from the microscope.
nr_peaks: int: Number of peaks to be extracted from the PCM (Default: 8)
nr_slices: int: Only applicable when running with 3D pictures and using ‘compres pic’ method in pairwisesingle.py. Determines the number of slices that are compressed together (compression in the z-direction). If None, all the slices are compressed

together. (Default: None)
nr_dim: int: If 3, the code will assume three dimensional data for the tile, where z is the first dimension and y and x the second and third. For any other value 2-dimensional data is assumed. (default: 2)

: dict: Contains key ‘P’ with a 1D numpy array containing pairwise alignment y and x coordinates (and z-coordinates when applicable) for each neighbouring pair of tiles, array will be 2 * len(contig_typles) for 2D data or 3 * len(contig_typles) for 3D data. Also contains key ‘covs’ with a 1D numpy array containing covariance for each pairwise alignment in ‘P’, ‘covs’ will be len(contig_typles).

pysmFISH.stitching_package.stitching.get_pairwise_input(ImageProperties, folder, tile_file, hyb_nr, gene='Nuclei', pre_proc_level='FilteredData', est_overlap=0.1, y_flip=False, nr_dim=2)[source]¶

Get the information necessary to do the pairwise allignment

Find the pairwise pars for an unknown stitching. Works best with a folder containing image with nuclei (DAPI staining)

folder: str: String representing the path of the folder containing the tile file and the yaml metadata file. Needs a trailing slash (‘/’).
tile_file: pointer: HDF5 file handle. Reference to the opened file containing the tiles.
hyb_nr: int: The number of the hybridization we are going to stitch. This will be used to navigate tile_file and find the correct tiles.
gene: str: The name of the gene we are going to stitch. This will be used to navigate tile_file and find the correct tiles. (Default: ‘Nuclei’)
pre_proc_level: str: The name of the pre processing group of the tiles we are going to stitch. This will be used to navigate tile_file and find the correct tiles. (Default: ‘Filtered’)
est_overlap: float: The fraction of two neighbours that should overlap, this is used to estimate the shape of the tile set and then overwritten by the actual average overlap according to the microscope coordinates. (default: 0.1)
y_flip: bool: The y_flip variable is designed for the cases where the microscope sequence is inverted in the y-direction. When set to True the y-coordinates will also be inverted before determining the tile set. (Default: False)
nr_dim: int: If 3, the code will assume three dimensional data for the tile, where z is the first dimension and y and x the second and third. For any other value 2-dimensional data is assumed. (Default: 2)

tiles: list: List of references to the the tiles in the hdf5 file tile_file.
contig_tuples: list: List of tuples. Each tuple is a tile pair. Tuples contain two tile indexes denoting these tiles are contingent to each other.
nr_pixels: int: Height and length of the tile in pixels, tile is assumed to be square.
z_count: int: The number of layers in one tile (size of the z-axis). Is 1 when nr_dim is not 3.
micData: object: MicroscopeData object. Contains coordinates of the tile corners as taken from the microscope.

pysmFISH.stitching_package.stitching.get_pairwise_input_npy(image_properties, converted_positions, hybridization, est_overlap, y_flip=False, nr_dim=2)[source]¶

Get the information necessary to do the pairwise allignment Modified version of the get_pairwise_input functions that work on .npy files and not on hdf5

Find the pairwise pairs for an unknown stitching.

image_properties: dict: Dictionary with the image details parsed from the Experimental_metadata.yaml file
converted_positions: dict: Dictionary with the coords of the images for all hybridization The coords are a list of floats
hybridization: str: Hybridization that will be processed (Ex. Hybridization2)
est_overlap: float: The fraction of two neighbours that should overlap, this is used to estimate the shape of the tile set and then overwritten by the actual average overlap according to the microscope coordinates. (default: 0.1)
y_flip: bool: The y_flip variable is designed for the cases where the microscope sequence is inverted in the y-direction. When set to True the y-coordinates will also be inverted before determining the tile set. (Default: False)
nr_dim: int: If 3, the code will assume three dimensional data for the tile, where z is the first dimension and y and x the second and third. For any other value 2-dimensional data is assumed. (Default: 2)

tiles: np.array: Array of int with the tiles number. -1 indicate an empty tile
contig_tuples: list: List of tuples. Each tuple is a tile pair. Tuples contain two tile indexes denoting these tiles are contingent to each other.
nr_pixels: int: Height and length of the tile in pixels, tile is assumed to be square.
z_count: int: The number of layers in one tile (size of the z-axis). Is 1 when nr_dim is not 3.
micData: object: MicroscopeData object. Contains coordinates of the tile corners as taken from the microscope.

pysmFISH.stitching_package.stitching.get_place_tile_input(folder, tiles, contig_tuples, micData, nr_pixels, z_count, alignment, data_name, nr_dim=2, save_alignment=True)[source]¶

Do the global alignment and get the shifted corner coordinates.

Calculates a shift in global coordinates for each tile (global alignment) and then applies these shifts to the corner coordinates of each tile and returns and saves these shifted corner coordinates.

This function produces a file with stitching data in folder called data_name, this file includes the corner coordinates which can be used to apply the stitching to another gene.

folder: str: String representing the path of the folder containing the tile file and the yaml metadata file. Needs a trailing slash (‘/’).
tiles: list: List of strings. List of references to the the tiles in the hdf5 file tile_file.
contig_tuples: list: List of tuples. Each tuple is a tile pair. Tuples contain two tile indexes denoting these tiles are contingent to each other.
micData: object: MicroscopeData object. Contains coordinates of the tile corners as taken from the microscope.
nr_pixels: int: Height and length of the tile in pixels, tile is assumed to be square.
z_count: int: The number of layers in one tile (size of the z-axis). Is 1 when nr_dim is not 3.
alignment: dict: Contains key ‘P’ with a 1D numpy array containing pairwise alignment y and x coordinates (and z-coordinates when applicable) for each neighbouring pair of tiles, array will be 2 * len(contig_typles) for 2D data or 3 * len(contig_typles) for 3D data. Also contains key ‘covs’ with a 1D numpy array containing covariance for each pairwise alignment in ‘P’, ‘covs’ will be len(contig_typles).
data_name: str: Name of the file containing the pickled stitching data.
nr_dim: int: If 3, the code will assume three dimensional data for the tile, where z is the first dimension and y and x the second and third. For any other value 2-dimensional data is assumed. (default: 2)
save_alignment: bool: When False only the stitching coordinates and microscope data will be saved. When True also the contigency tuples and pairwise alignment will be saved (this is necessary if we want to refine the stitching later). (Default: True)

joining: dict: Contains keys corner_list and final_image_shape. Corner_list is a list of list, each list is a pair of a tile index (int) and it’s tile’s shifted coordinates in the final image (numpy array containing floats). Final_image_shape is a tuple of size 2 or 3 depending on the numer of dimensions and contains ints.

pysmFISH.stitching_package.stitching.get_place_tile_input_apply(folder, tile_file, hyb_nr, data_name, gene='Nuclei', pre_proc_level='Filtered', nr_dim=2, check_pairwise=False)[source]¶

Get the data needed to apply stitching to another gene

folder: str: String representing the path of the folder containing the tile file, the stitching data file the yaml metadata file. Needs a trailing slash (‘/’).
tile_file: pointer: HDF5 file handle. Reference to the opened file containing the tiles.
hyb_nr: int: The number of the hybridization we are going to stitch. This will be used to navigate tile_file and find the correct tiles.
data_name: str: Name of the file containing the pickled stitching data.
gene: str: The name of the gene we are going to stitch. This will be used to navigate tile_file and find the correct tiles. (Default: ‘Nuclei’)
pre_proc_level: str: The name of the pre processing group of the tiles we are going to stitch. This will be used to navigate tile_file and find the correct tiles. (Default: ‘Filtered’)
nr_dim: int: If 3, the code will assume three dimensional data for the tile, where z is the first dimension and y and x the second and third. For any other value 2-dimensional data is assumed. (Default: 2)
check_pairwise: bool: If True the contig_tuples array is assumed to be in the pickled data file and will be returned. (Default: False)

joining: dict: Taken from the stitching data file. Contains keys corner_list and final_image_shape. Corner_list is a list of list, each list is a pair of an image number (int) and it’s coordinates (numpy array containing floats). Final_image_shape is a tuple of size 2 or 3 depending on the numer of dimensions and contains ints.
tiles: list: List of references to the the tiles in the hdf5 file tile_file.
nr_pixels: int: Height and length of the tile in pixels, tile is assumed to be square.
z_count: int: The number of layers in one tile (size of the z-axis). Is 1 when nr_dim is not 3.
micData: object: MicroscopeData object. Taken from the pickled stitching data. Contains coordinates of the tile corners as taken from the microscope.
contig_tuples: list: Only returned if check_pairwise == True. list of tuples. Taken from the pickled stitching data. Each tuple is a tile pair. Tuples contain two tile indexes denoting these tiles are contingent to each other.

pysmFISH.stitching_package.stitching.get_place_tile_input_apply_npy(hyb_dir, stitched_reference_files_dir, data_name, image_properties, nr_dim=2)[source]¶

Modified version of the get_place_tile_input_apply Get the data needed to apply stitching to another gene

hyb_dir: str: String representing the path of the folder containing the tile file, the stitching data file the yaml metadata file.
stitched_reference_files_dir: str: String representing the path of the folder containing the registered data.
data_name: str: Name of the file containing the pickled stitching data.
image_properties: dict: Dictionary with the image details parsed from the Experimental_metadata.yaml file
nr_dim: int: If 3, the code will assume three dimensional data for the tile, where z is the first dimension and y and x the second and third. For any other value 2-dimensional data is assumed. (Default: 2)

joining: dict: Taken from the stitching data file. Contains keys corner_list and final_image_shape. Corner_list is a list of list, each list is a pair of an image number (int) and it’s coordinates (numpy array containing floats). Final_image_shape is a tuple of size 2 or 3 depending on the numer of dimensions and contains ints.
tiles: list: List of strings. List of references to the the tiles in the hdf5 file tile_file.
nr_pixels: int: Height and length of the tile in pixels, tile is assumed to be square.
z_count: int: The number of layers in one tile (size of the z-axis). Is 1 when nr_dim is not 3.
micData: object: MicroscopeData object. Taken from the pickled stitching data. Contains coordinates of the tile corners as taken from the microscope.

pysmFISH.stitching_package.stitching.get_refine_pairwise_input(folder, tile_file, hyb_nr, data_name, gene='Nuclei', pre_proc_level='Filtered', nr_dim=2)[source]¶

Get the data needed to refine stitching with another gene

folder: str: String representing the path of the folder containing the tile file, the stitching data file the yaml metadata file. Needs a trailing slash (‘/’).
tile_file: pointer: HDF5 file handle. Reference to the opened file containing the tiles.
hyb_nr: int: The number of the hybridization we are going to stitch. This will be used to navigate tile_file and find the correct tiles.
data_name: str: Name of the file containing the pickled stitching data.
gene: str: The name of the gene we are going to stitch. This will be used to navigate tile_file and find the correct tiles. (Default: ‘Nuclei’)
pre_proc_level: str: The name of the pre processing group of the tiles we are going to stitch. This will be used to navigate tile_file and find the correct tiles. (Default: ‘Filtered’)
nr_dim: int: If 3, the code will assume three dimensional data for the tile, where z is the first dimension and y and x the second and third. For any other value 2-dimensional data is assumed. (Default: 2)

tiles: list: List of references to the the tiles in the hdf5 file tile_file.
contig_tuples: list: List of tuples. Each tuple is a tile pair. Tuples contain two tile indexes denoting these tiles are contingent to each other.
nr_pixels: int: Height and length of the tile in pixels, tile is assumed to be square.
z_count: int: The number of layers in one tile (size of the z-axis). Is 1 when nr_dim is not 3.
micData: object: MicroscopeData object. Contains coordinates of the tile corners as taken from the microscope.

pysmFISH.stitching_package.stitching.plot_final_image(im_file_name, joining, hyb_nr=1, gene='Nuclei', fig_name='final image', shrink_image=False, block=True)[source]¶

Displays the high quality final image in a plot window.

Takes a lot of working memory for full sized images. When plt_available is false this function does nothing and returns None.

im_file_name: str: Filename of the hdf5 file, containing the final image.
fig_name: str: Name of the plotting window (default: “final image”).
shrink_image: bool: Turn on shrink_image to reduce display quality and memory usage. (Default: False)
block: bool: Plot blocks the running program untill the plotting window is closed if true. Turn off block to make the code continue untill the next call of plt.show(block=True) before displaying the image. (default: True)

pysmFISH.stitching_package.stitching.refine_pairwise_alignments(tiles, tile_file, contig_tuples, alignment_old, micData=None, nr_peaks=8, nr_dim=2)[source]¶

Calculate the pairwise transition

Calculates pairwise transition for each neighbouring pair of tiles.

tiles: np.array: Array of tiles, a tile should be a 2d np.array representing a picture
contig_tuples: list: List of tuples denoting which tiles are contingent to each other.
micData: object: MicroscopeData object containing coordinates (default None)
nr_peaks: int: nr of peaks to be extracted from the PCM (default 8)
nr_dim: int: If 3, the code will assume three dimensional data for the tile, where z is the first dimension and y and x the second and third. For any other value 2-dimensional data is assumed. (default: 2)

: dict: Contains key ‘P’ with a 1D numpy array containing pairwise alignment y and x coordinates (and z-coordinates when applicable) for each neighbouring pair of tiles, array will be 2 * len(contig_typles) for 2D data or 3 * len(contig_typles) for 3D data. Also contains key ‘covs’ with a 1D numpy array containing covariance for each pairwise alignment in ‘P’, ‘covs’ will be len(contig_typles).

pysmFISH.stitching_package.stitching.save_as_tiff(data_file, hyb_nr, gene, location_image, pre_proc_level='StitchedImage', mode='both')[source]¶

Save the results as a tiff image for visual inspection.

data_file: pointer: HDF5 file handle. HDF5 file containing the final image.
gene: str: The name of the gene we stitched. This will be used to navigate data_file and find the correct final picture.
hyb_nr: int: The number of the hybridization we have stitched.This will be used to navigate data_file and find the correct final picture.
location_image: str: Full path to the file where the tiff file will be saved (extension not necessary).
pre_proc_level: str: The name of the pre processing group of the tiles we are going to stitch. Normally this will be ‘StitchedImage’, but when the final image is found in another datagroup it may be changed. This will be used to navigate data_file and find the correct final image. (Default: ‘StitchedImage’)
mode: str: Mode determines what color, quality and how many images are saved. Possible values for mode: save_ubyte, save_float, save_rgb. If another or no value is given the image is saved as is and a as a low quality copy (pixel depth 8 bits) (Default: ‘both’)