Skip to content

sync branch #41

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
guillaumelegoc merged 14 commits into from
May 28, 2025
Merged

sync branch #41

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
14 commits
Select commit Hold shift + click to select a range
2b7f2ba
image white background
guillaumelegoc May 23, 2025
f5088f2
Better example
guillaumelegoc May 23, 2025
e95e2c7
Update cells_distributions.py, density_map.py, fibers_coverage.ipynb,…
guillaumelegoc May 23, 2025
70c9522
allow for missing file so that it can be corrected later
guillaumelegoc May 26, 2025
eb93e00
self-runnable examples as scripts (from existing notebooks)
guillaumelegoc May 26, 2025
c026e00
wip self-running segmentation example
guillaumelegoc May 26, 2025
b50d164
Update segment_images.py, example-seg.tar.gz, and animalid0_001_Proba…
guillaumelegoc May 26, 2025
13666c6
Update segment_images.py
guillaumelegoc May 26, 2025
dc9aaee
Merge pull request #35 from TeamNCMC/doc-white-bg
guillaumelegoc May 26, 2025
c9d230d
Merge pull request #36 from TeamNCMC/fix-batch-example
guillaumelegoc May 26, 2025
6ca96c7
Merge pull request #37 from TeamNCMC/duplicate-examples
guillaumelegoc May 26, 2025
6bda783
Merge pull request #38 from TeamNCMC/include-data-seg
guillaumelegoc May 26, 2025
9f80c29
Merge pull request #40 from TeamNCMC/duplicate-examples
guillaumelegoc May 28, 2025
fd10cd2
Update pyproject.toml
guillaumelegoc May 28, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
8 changes: 5 additions & 3 deletions cuisto/config.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4,6 +4,7 @@

"""

import os
import tomllib
import warnings

Expand Down Expand Up @@ -62,9 +63,10 @@ def __init__(self, config_file):
def get_blacklist(self):
"""Wraps cuisto.utils.get_blacklist."""

self.atlas["blacklist"] = utils.get_blacklist(
self.files["blacklist"], self.bg_atlas
)
if os.path.isfile(self.files["blacklist"]):
self.atlas["blacklist"] = utils.get_blacklist(
self.files["blacklist"], self.bg_atlas
)

def get_leaves_list(self):
"""Wraps utils.get_leaves_list."""
Expand Down
128 changes: 54 additions & 74 deletions docs/images/cuisto-showcase.svg
View file
Open in desktop
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
4 changes: 2 additions & 2 deletions examples/batch_process_animals.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -37,8 +37,8 @@
# full path to where are the TSV files and {animal}_detections folders
# dl_example = False
# animals = ("animalid0", "animalid1")
# input_dir = /path/to/your/data
# config_file = /path/to/your/config.toml
# input_dir = "~/.cuisto/example"
# config_file = "~/.cuisto/config_multi.toml"

# Download example data into cuisto default folder. Remove this block if using your data
if dl_example:
Expand Down
97 changes: 97 additions & 0 deletions examples/cells_distributions.py
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,97 @@
# This script shows how to load data exported from QuPath, compute metrics and display
# them, according to the configuration file. This is meant for a single-animal.
#
# There are some conventions that need to be met in the QuPath project so that the
# measurements are usable with `cuisto`:
# + Objects' classifications must be derived, eg. be in the form "something: else". The
# primary classification ("something") will be refered to "object_type" and the
# secondary classification ("else") to "channel" in the configuration file.
# + Only one "object_type" can be processed at once, but supports any numbers of
# channels.
# + Annotations (brain regions) must have properly formatted measurements. For punctual
# objects, it would be the count. Run the "add_regions_count.groovy" script to add
# them. The measurements names must be in the form "something: else name", for
# instance, "something: else Count". "name" is refered to "base_measurement" in the
# configuration file.
#
# The data was generated from QuPath with stardist cell detection followed by a pixel
# classifier "Classify" function on toy data.
#
# If you cloned the repository, this example should be ready to run.
#
# Find this script as a notebook in the documentation :
# https://teamncmc.github.io/cuisto/demo_notebooks/cells_distributions.html

from pathlib import Path

import matplotlib.pyplot as plt
import pandas as pd

import cuisto

# Full path to your configuration file, edited according to your need beforehand
wdir = Path(__file__).parent.parent / "resources"
config_file = wdir / "demo_config_cells.toml"

# - Files
# animal identifier
animal = "animalid0"
# set the full path to the annotations tsv file from QuPath
annotations_file = wdir / "cells_measurements_annotations.tsv"
# set the full path to the detections tsv file from QuPath
detections_file = wdir / "cells_measurements_detections.tsv"

# get configuration
cfg = cuisto.config.Config(config_file)
# update configuration file paths (so that this example can self-run)
cfg.files["blacklist"] = wdir / "demo_atlas_blacklist_brain.toml"
cfg.files["fusion"] = wdir / "demo_atlas_fusion_brain.toml"
cfg.files["infos"] = wdir / "demo_info_cells.toml"
cfg.get_blacklist()

# read data
df_annotations = pd.read_csv(annotations_file, index_col="Object ID", sep="\t")
df_detections = pd.read_csv(detections_file, index_col="Object ID", sep="\t")

# remove annotations that are not brain regions
df_annotations = df_annotations[df_annotations["Classification"] != "Region*"]
df_annotations = df_annotations[df_annotations["ROI"] != "Rectangle"]

# convert atlas coordinates from mm to microns
df_detections[["Atlas_X", "Atlas_Y", "Atlas_Z"]] = df_detections[
["Atlas_X", "Atlas_Y", "Atlas_Z"]
].multiply(1000)

# have a look
print(df_annotations.head())
print(df_detections.head())

# get distributions per regions, spatial distributions and coordinates
df_regions, dfs_distributions, df_coordinates = cuisto.process.process_animal(
animal, df_annotations, df_detections, cfg, compute_distributions=True
)

# have a look
print(df_regions.head())
print(df_coordinates.head())

# plot distributions per regions
figs_regions = cuisto.display.plot_regions(df_regions, cfg)
# specify which regions to plot
# figs_regions = cuisto.display.plot_regions(df_regions, cfg, names_list=["GRN", "IRN", "MDRNv"])

# save as svg
# figs_regions[0].savefig(r"C:\Users\glegoc\Downloads\regions_count.svg")
# figs_regions[1].savefig(r"C:\Users\glegoc\Downloads\regions_density.svg")

# plot 1D distributions
fig_distrib = cuisto.display.plot_1D_distributions(
dfs_distributions, cfg, df_coordinates=df_coordinates
)
# If there were several `animal` in the measurement file, it would be displayed as
# mean +/- sem instead.

# plot heatmap (all types of cells pooled)
fig_heatmap = cuisto.display.plot_2D_distributions(df_coordinates, cfg)

plt.show()
167 changes: 167 additions & 0 deletions examples/density_map.py
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,167 @@
# Draw 2D heatmaps as density isolines.
#
# This script does not actually use `cuisto` and relies only on
# brainglobe-heatmap (https://brainglobe.info/documentation/brainglobe-heatmap/index.html)
# to extract brain structures outlines.
#
# Only the detections measurements with atlas coordinates exported from QuPath are used.
#
# You need to select the range of data to be used, the regions outlines will be
# extracted at the centroid of that range. Therefore, a range that is too large will be
# misleading and irrelevant.
#
# If you cloned the repository, this example should be ready to run.
#
# Find this script as a notebook in the documentation :
# https://teamncmc.github.io/cuisto/demo_notebooks/density_map.html

from pathlib import Path

import brainglobe_heatmap as bgh
import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
import seaborn as sns

# path to the exported measurements from QuPath
wdir = Path(__file__).parent.parent / "resources"
filename = wdir / "cells_measurements_detections.tsv"

# Settings
# atlas to use
atlas_name = "allen_mouse_10um"
# brain regions whose outlines will be plotted
regions = ["root", "CB", "MY", "GRN", "IRN"]
# range to include, in Allen coordinates, in microns
ap_lims = [9800, 10000] # lims : [0, 13200] for coronal
ml_lims = [5600, 5800] # lims : [0, 11400] for sagittal
dv_lims = [3900, 4100] # lims : [0, 8000] for top
# number of isolines
nlevels = 5
# color mapping between classification and matplotlib color
palette = {"Cells: marker-": "#d8782f", "Cells: marker+": "#8ccb73"}

df = pd.read_csv(filename, sep="\t")
print(df.head())

# Here we can filter out classifications we don't wan't to display.
# select objects
# df = df[df["Classification"] == "example: classification"]

# get outline coordinates in coronal (=frontal) orientation
coords_coronal = bgh.get_structures_slice_coords(
regions,
orientation="frontal",
atlas_name=atlas_name,
position=(np.mean(ap_lims), 0, 0),
)
# get outline coordinates in sagittal orientation
coords_sagittal = bgh.get_structures_slice_coords(
regions,
orientation="sagittal",
atlas_name=atlas_name,
position=(0, 0, np.mean(ml_lims)),
)
# get outline coordinates in top (=horizontal) orientation
coords_top = bgh.get_structures_slice_coords(
regions,
orientation="horizontal",
atlas_name=atlas_name,
position=(0, np.mean(dv_lims), 0),
)

# Coronal projection
# select objects within the rostro-caudal range
df_coronal = df[
(df["Atlas_X"] >= ap_lims[0] / 1000) & (df["Atlas_X"] <= ap_lims[1] / 1000)
]

plt.figure()

for struct_name, contours in coords_coronal.items():
for cont in contours:
plt.fill(cont[:, 0] / 1000, cont[:, 1] / 1000, lw=1, fc="none", ec="k")

# see https://seaborn.pydata.org/generated/seaborn.kdeplot.html to customize
ax = sns.kdeplot(
df_coronal,
x="Atlas_Z",
y="Atlas_Y",
hue="Classification",
levels=nlevels,
common_norm=False,
palette=palette,
)
ax.invert_yaxis()
sns.despine(left=True, bottom=True)
plt.axis("equal")
plt.xlabel(None)
plt.ylabel(None)
plt.xticks([])
plt.yticks([])
plt.plot([2, 3], [8, 8], "k", linewidth=3)
plt.text(2, 7.9, "1 mm")

# Sagittal projection
# select objects within the medio-lateral range
df_sagittal = df[
(df["Atlas_Z"] >= ml_lims[0] / 1000) & (df["Atlas_Z"] <= ml_lims[1] / 1000)
]

plt.figure()

for struct_name, contours in coords_sagittal.items():
for cont in contours:
plt.fill(cont[:, 0] / 1000, cont[:, 1] / 1000, lw=1, fc="none", ec="k")

# see https://seaborn.pydata.org/generated/seaborn.kdeplot.html to customize
ax = sns.kdeplot(
df_sagittal,
x="Atlas_X",
y="Atlas_Y",
hue="Classification",
levels=nlevels,
common_norm=False,
palette=palette,
)
ax.invert_yaxis()
sns.despine(left=True, bottom=True)
plt.axis("equal")
plt.xlabel(None)
plt.ylabel(None)
plt.xticks([])
plt.yticks([])
plt.plot([2, 3], [7.1, 7.1], "k", linewidth=3)
plt.text(2, 7, "1 mm")

# Top projection
# select objects within the dorso-ventral range
df_top = df[(df["Atlas_Y"] >= dv_lims[0] / 1000) & (df["Atlas_Y"] <= dv_lims[1] / 1000)]

plt.figure()

for struct_name, contours in coords_top.items():
for cont in contours:
plt.fill(-cont[:, 0] / 1000, cont[:, 1] / 1000, lw=1, fc="none", ec="k")

# see https://seaborn.pydata.org/generated/seaborn.kdeplot.html to customize
ax = sns.kdeplot(
df_top,
x="Atlas_Z",
y="Atlas_X",
hue="Classification",
levels=nlevels,
common_norm=False,
palette=palette,
)
ax.invert_yaxis()
sns.despine(left=True, bottom=True)
plt.axis("equal")
plt.xlabel(None)
plt.ylabel(None)
plt.xticks([])
plt.yticks([])
plt.plot([0.5, 1.5], [0.5, 0.5], "k", linewidth=3)
plt.text(0.5, 0.4, "1 mm")

plt.show()
78 changes: 78 additions & 0 deletions examples/fibers_coverage.py
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,78 @@
# Plot regions coverage percentage in the spinal cord.
#
# This showcases that any brainglobe atlases should be supported.
#
# Here we're going to quantify the percentage of area of each spinal cord regions
# innervated by axons.
#
# The "area µm^2" measurement for each annotations can be created in QuPath with a pixel
# classifier, using the Measure button.
#
# We're going to consider that the "area µm^2" measurement generated by the pixel
# classifier is an object count.
# `cuisto` computes a density, which is the count in each region divided by its aera.
# Therefore, in this case, it will be actually the fraction of area covered by fibers
# in a given color.
#
# The data was generated using QuPath with a pixel classifier on toy data.
#
# If you cloned the repository, this example should be ready to run.
#
# Find this script as a notebook in the documentation :
# https://teamncmc.github.io/cuisto/demo_notebooks/fibers_coverage.html

from pathlib import Path

import matplotlib.pyplot as plt
import pandas as pd

import cuisto

# Full path to your configuration file, edited according to your need beforehand
wdir = Path(__file__).parent.parent / "resources"
config_file = wdir / "demo_config_fibers.toml"

# - Files
# not important if only one animal
animal = "animalid1-SC"
# set the full path to the annotations tsv file from QuPath
annotations_file = wdir / "fibers_measurements_annotations.tsv"

# get configuration
cfg = cuisto.config.Config(config_file)
# update configuration file paths (so that this example can self-run)
cfg.files["blacklist"] = wdir / "demo_atlas_blacklist_cord.toml"
cfg.files["fusion"] = wdir / "demo_atlas_fusion_cord.toml"
cfg.get_blacklist()

# read data
df_annotations = pd.read_csv(annotations_file, index_col="Object ID", sep="\t")
df_detections = pd.DataFrame() # empty DataFrame

# remove annotations that are not brain regions
df_annotations = df_annotations[df_annotations["Classification"] != "Region*"]
df_annotations = df_annotations[df_annotations["ROI"] != "Rectangle"]

# have a look
print(df_annotations.head())

# get distributions per regions, spatial distributions and coordinates
df_regions, dfs_distributions, df_coordinates = cuisto.process.process_animal(
animal, df_annotations, df_detections, cfg, compute_distributions=False
)

# convert the "density µm^-2" column, which is actually the coverage fraction, to a percentage
df_regions["density µm^-2"] = df_regions["density µm^-2"] * 100

# have a look
print(df_regions.head())

# plot distributions per regions
fig_regions = cuisto.display.plot_regions(df_regions, cfg)
# specify which regions to plot
# fig_regions = hq.display.plot_regions(df_regions, cfg, names_list=["Rh9", "Sr9", "8Sp"])

# save as svg
# fig_regions[0].savefig(r"C:\Users\glegoc\Downloads\nice_figure.svg")

plt.show()
Loading
Loading