import os
import tifffile as tif
import numpy as np
import napari
import pandas as pd
import matplotlib.pyplot as plt
import warnings

# change settings to show all columns in a dataframe
pd.set_option('display.max_columns', None)

# ignore napari warnings
with warnings.catch_warnings():
    warnings.simplefilter("ignore")
    # create napari viewer instance
    viewer = napari.Viewer() 

Note: notebook 4_quantitative_measurements needs to be run again for a different compound before starting this one.

path = "/Users/laura/projects/Bio-image_analysis_school_ScadsAI/prepared_dataset"

nuclei_labels_path = os.path.join(path, "labels_nuclei")
actin_labels_path = os.path.join(path, "labels_actin")
tubulin_labels_path = os.path.join(path, "labels_tubulin")

os.listdir(path)

['aphidicolin_timelapse.tif',
 '.DS_Store',
 'nocodazole_timelapse.tif',
 'AZ-I_timelapse.tif',
 'taxol_timelapse.tif',
 'labels_nuclei',
 'latrunculin B_timelapse.tif',
 'epothilone B_timelapse.tif',
 'monastrol_timelapse.tif',
 'AZ-H_timelapse.tif',
 'DMSO_timelapse.tif',
 'colchicine_timelapse.tif',
 'doxorubicin_timelapse.tif',
 'cytochalasin B_timelapse.tif',
 'labels_tubulin',
 'labels_actin',
 'AZ-A_timelapse.tif']

filename1 = "nocodazole_timelapse"
filename2 = "latrunculin B_timelapse"
filename_control= "DMSO_timelapse"

img1 = tif.imread(os.path.join(path, f'{filename1}.tif'))
img2 = tif.imread(os.path.join(path, f'{filename2}.tif'))

img_c = tif.imread(os.path.join(path, f'{filename_control}.tif'))

# load nuclei and actin segmentation
labels_nuclei1 = tif.imread(os.path.join(nuclei_labels_path, f'{filename1}_labels_dapi.tif'))
labels_actin1 = tif.imread(os.path.join(actin_labels_path, f'{filename1}_labels_actin.tif'))
labels_tubulin1 = tif.imread(os.path.join(tubulin_labels_path, f'{filename1}_labels_tubulin.tif'))

labels_nuclei2 = tif.imread(os.path.join(nuclei_labels_path, f'{filename2}_labels_dapi.tif'))
labels_actin2 = tif.imread(os.path.join(actin_labels_path, f'{filename2}_labels_actin.tif'))
labels_tubulin2 = tif.imread(os.path.join(tubulin_labels_path, f'{filename2}_labels_tubulin.tif'))

labels_nuclei_c = tif.imread(os.path.join(nuclei_labels_path, f'{filename_control}_labels_dapi.tif'))
labels_actin_c = tif.imread(os.path.join(actin_labels_path, f'{filename_control}_labels_actin.tif'))
labels_tubulin_c = tif.imread(os.path.join(tubulin_labels_path, f'{filename_control}_labels_tubulin.tif'))

img1 = img1[:, np.newaxis, :, :, :]

# do the same for all label images
labels_actin1 = labels_actin1[:, np.newaxis, :, :]
labels_nuclei1 = labels_nuclei1[:, np.newaxis, :, :]
labels_tubulin1 = labels_tubulin1[:, np.newaxis, :, :]

img2 = img2[:, np.newaxis, :, :, :]

# do the same for all label images
labels_actin2 = labels_actin2[:, np.newaxis, :, :]
labels_nuclei2 = labels_nuclei2[:, np.newaxis, :, :]
labels_tubulin2 = labels_tubulin2[:, np.newaxis, :, :]

img_c = img_c[:, np.newaxis, :, :, :]

# do the same for all label images
labels_actin_c = labels_actin_c[:, np.newaxis, :, :]
labels_nuclei_c = labels_nuclei_c[:, np.newaxis, :, :]
labels_tubulin_c = labels_tubulin_c[:, np.newaxis, :, :]

# since with napari-clusters-plotter we can use only image, we will combine both images into one timelapse, 
# where in timepoints 0-7 we have compound 1, in 8-15 timepoints we have compound 2, and in 16th control DMSO
print(f"Shape of the first image: {img1.shape}")
print(f"Shape of the second image: {img2.shape}")
print(f"Shape of the control image: {img_c.shape}")

# the shapes of both images need to be the same to combine them like this:
combined_timelapse = np.concatenate((img1, img2, img_c), axis=0)

Shape of the first image: (8, 1, 1024, 1280, 3)
Shape of the second image: (8, 1, 1024, 1280, 3)
Shape of the control image: (1, 1, 1024, 1280, 3)

# viewer.add_image(combined_timelapse) # and then in the viewer right click on the layer - split RGB or:
viewer.add_image(
    combined_timelapse,
    name=["tubulin", "actin", "nuclei"],
    colormap=["magenta", "green", "blue"],
    channel_axis=4)

[<Image layer 'tubulin' at 0x28f5bfe50>,
 <Image layer 'actin' at 0x1721b9fa0>,
 <Image layer 'nuclei' at 0x1721f9c70>]

# do the same for label images
combined_actin_labels = np.concatenate((labels_actin1, labels_actin2, labels_actin_c), axis=0)
combined_tubulin_labels = np.concatenate((labels_tubulin1, labels_tubulin2, labels_tubulin_c), axis=0)
combined_nuclei_labels = np.concatenate((labels_nuclei1, labels_nuclei2, labels_nuclei_c), axis=0)

# add segmentation layers to the viewer
viewer.add_labels(combined_actin_labels, name="actin_segmented")
viewer.add_labels(combined_nuclei_labels, name="nuclei_segmented")
viewer.add_labels(combined_tubulin_labels, name="tubulin_segmented")

<Labels layer 'tubulin_segmented' at 0x2f332ac40>

# load corresponding measurements csv
df1 = pd.read_csv(f'{filename1}_measurements.csv')
df2 = pd.read_csv(f'{filename2}_measurements.csv')
df_c = pd.read_csv(f'{filename_control}_measurements.csv')

# combine both measurements tables into one
offset = df1['frame'].max() + 1

# adjust frame column for the second compound
df2['frame'] += offset

# new offset for the control image, which is the last one
offset = df2['frame'].max() + 1
df_c['frame'] += offset

combined_df = pd.concat([df1, df2, df_c])

combined_df

	frame	label	area_dapi	equivalent_diameter_dapi	mean_intensity_dapi	min_intensity_dapi	max_intensity_dapi	perimeter_dapi	perimeter_crofton_dapi	extent_dapi	solidity_dapi	feret_diameter_max_dapi	major_axis_length_dapi	minor_axis_length_dapi	eccentricity_dapi	orientation_dapi	area_actin	equivalent_diameter_actin	mean_intensity_actin	min_intensity_actin	max_intensity_actin	perimeter_actin	perimeter_crofton_actin	extent_actin	solidity_actin	feret_diameter_max_actin	major_axis_length_actin	minor_axis_length_actin	eccentricity_actin	orientation_actin	area_tubulin	equivalent_diameter_tubulin	mean_intensity_tubulin	min_intensity_tubulin	max_intensity_tubulin	perimeter_tubulin	perimeter_crofton_tubulin	extent_tubulin	solidity_tubulin	feret_diameter_max_tubulin	major_axis_length_tubulin	minor_axis_length_tubulin	eccentricity_tubulin	orientation_tubulin
0	0	1	51.0	8.058239	0.167979	0.083643	0.254647	34.485281	35.375614	0.750000	0.944444	17.000000	16.699044	4.066449	0.969897	-1.564183	883.0	33.530143	0.119379	0.081159	0.192271	146.876154	144.414011	0.617051	0.899185	53.460266	50.832770	24.742622	0.873544	-1.468765	1478.0	43.380273	0.143015	0.029722	0.366054	206.917785	201.337046	0.565633	0.856812	70.519501	66.715717	34.099459	0.859512	1.318986
1	0	2	649.0	28.745999	0.165379	0.115242	0.210037	101.112698	98.542366	0.780048	0.941945	35.341194	34.562224	25.219399	0.683788	1.438942	1389.0	42.053891	0.155570	0.106280	0.219324	183.710678	177.242860	0.665230	0.871940	59.203040	56.035253	34.286886	0.790950	1.459215	2416.0	55.463021	0.100016	0.029331	0.238952	252.409163	242.765807	0.567136	0.865330	84.059503	79.006359	41.878135	0.847960	-0.483112
2	0	3	965.0	35.052477	0.178177	0.087361	0.250929	116.811183	113.425463	0.724474	0.971803	41.880783	41.374456	29.809831	0.693467	-0.855667	5343.0	82.479809	0.152169	0.045411	0.453140	557.830519	546.056318	0.466190	0.676244	159.840546	156.791723	56.600349	0.932569	-1.338031	3355.0	65.358386	0.140226	0.029331	0.383653	304.758405	294.881169	0.513468	0.862690	114.908659	94.147899	49.308249	0.851883	-0.998811
3	0	4	681.0	29.446156	0.190088	0.100372	0.327138	105.254834	102.469357	0.709375	0.935440	34.985711	30.833248	30.070002	0.221123	0.357786	1723.0	46.837931	0.192738	0.045411	0.591304	342.800036	328.820739	0.396457	0.471151	112.946890	111.582379	36.281730	0.945660	-1.458035	952.0	34.815572	0.165720	0.037544	0.412202	140.669048	136.044137	0.689855	0.878229	49.040799	46.718906	29.029679	0.783518	1.430018
4	0	5	176.0	14.969641	0.143112	0.117100	0.171004	56.627417	56.367675	0.666667	0.956522	24.083189	22.479143	11.103994	0.869480	-1.503134	468.0	24.410574	0.100213	0.079227	0.142995	98.805087	99.787840	0.609375	0.858716	34.928498	28.834912	22.419438	0.628870	-1.254344	1678.0	46.222245	0.098445	0.029331	0.338287	192.503571	186.168095	0.747772	0.853510	60.207973	49.546863	46.315471	0.355224	0.940611
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
218	16	219	612.0	27.914559	0.228325	0.162602	0.294309	95.941125	93.639408	0.809524	0.977636	36.674242	36.376578	22.083566	0.794639	1.503317	5728.0	85.399743	0.059505	0.033823	0.086790	399.385822	383.482865	0.623965	0.863952	153.209008	138.180600	57.224254	0.910219	-1.561466	5159.0	81.047164	0.127046	0.029626	0.507514	412.421356	399.897290	0.630993	0.861413	146.030819	131.954569	54.136550	0.911966	-1.549621
219	16	220	501.0	25.256544	0.168067	0.123577	0.217886	88.041631	86.150217	0.824013	0.972816	32.984845	32.388596	20.667636	0.769941	-1.563516	2712.0	58.762451	0.071224	0.033823	0.180600	282.929978	271.112306	0.623448	0.768054	88.842557	79.161028	52.439570	0.749113	1.341769	2522.0	56.666658	0.135754	0.029626	0.385144	274.237590	265.193911	0.596217	0.795333	96.664368	81.256961	46.949464	0.816185	1.271864
220	16	221	260.0	18.194567	0.138030	0.099187	0.175610	72.627417	71.536626	0.738636	0.966543	32.062439	31.240097	11.340435	0.931786	-1.568107	1604.0	45.191550	0.071948	0.034461	0.105935	210.124892	200.551648	0.668333	0.878423	75.239617	71.259864	32.045703	0.893179	1.474066	2334.0	54.513678	0.079832	0.029626	0.173036	279.166522	272.778387	0.736742	0.861891	101.202767	107.910973	29.962604	0.960679	-1.497768
221	16	222	172.0	14.798554	0.167338	0.087805	0.227642	65.935029	65.388194	0.792627	0.955556	31.016125	31.027703	7.395085	0.971182	-1.560152	2313.0	54.267882	0.089133	0.033823	0.176771	368.279221	353.010209	0.590805	0.642322	135.181360	151.749307	30.689377	0.979337	1.554727	1740.0	47.068427	0.124112	0.029626	0.381280	315.752309	304.552333	0.491803	0.602285	123.320720	131.987967	26.355545	0.979861	-1.521874
222	16	223	43.0	7.399277	0.149783	0.086179	0.206504	30.485281	31.583376	0.716667	0.934783	15.000000	14.664421	3.934382	0.963337	1.563685	1219.0	39.396434	0.077260	0.033823	0.130823	167.982756	161.939156	0.609500	0.863314	51.224994	44.378772	39.890167	0.438242	-1.499674	1269.0	40.196281	0.084577	0.029626	0.267926	176.639610	180.939813	0.666492	0.891152	56.142675	52.837074	33.841353	0.767971	1.503483

2739 rows × 44 columns

# add measurements table to the properties of the tubulin labels layer
viewer.layers['tubulin_segmented'].properties = combined_df

Selected labels layer: tubulin_segmented
Selected measurements: ['area_dapi', 'equivalent_diameter_dapi', 'mean_intensity_dapi', 'min_intensity_dapi', 'max_intensity_dapi', 'perimeter_dapi', 'perimeter_crofton_dapi', 'extent_dapi', 'solidity_dapi', 'feret_diameter_max_dapi', 'major_axis_length_dapi', 'minor_axis_length_dapi', 'eccentricity_dapi', 'orientation_dapi', 'area_actin', 'equivalent_diameter_actin', 'mean_intensity_actin', 'min_intensity_actin', 'max_intensity_actin', 'perimeter_actin', 'perimeter_crofton_actin', 'extent_actin', 'solidity_actin', 'feret_diameter_max_actin', 'major_axis_length_actin', 'minor_axis_length_actin', 'eccentricity_actin', 'orientation_actin', 'area_tubulin', 'equivalent_diameter_tubulin', 'mean_intensity_tubulin', 'min_intensity_tubulin', 'max_intensity_tubulin', 'perimeter_tubulin', 'perimeter_crofton_tubulin', 'extent_tubulin', 'solidity_tubulin', 'feret_diameter_max_tubulin', 'major_axis_length_tubulin', 'minor_axis_length_tubulin', 'eccentricity_tubulin', 'orientation_tubulin']

/Users/laura/micromamba/envs/devbio-napari-env/lib/python3.9/site-packages/umap/umap_.py:1943: UserWarning: n_jobs value -1 overridden to 1 by setting random_state. Use no seed for parallelism.
  warn(f"n_jobs value {self.n_jobs} overridden to 1 by setting random_state. Use no seed for parallelism.")

Napari status bar display of label properties disabled because https://github.com/napari/napari/issues/5417 and https://github.com/napari/napari/issues/4342
Dimensionality reduction finished

/Users/laura/micromamba/envs/devbio-napari-env/lib/python3.9/site-packages/napari/components/viewer_model.py:10: FutureWarning: Labels.color is deprecated since 0.4.19 and will be removed in 0.5.0, please set Labels.colormap directly with an instance of napari.utils.colormaps.DirectLabelColormap instead.
  TYPE_CHECKING,
/Users/laura/micromamba/envs/devbio-napari-env/lib/python3.9/site-packages/napari/utils/_proxies.py:144: FutureWarning: Labels.color is deprecated since 0.4.19 and will be removed in 0.5.0, please set Labels.colormap directly with an instance of napari.utils.colormaps.DirectLabelColormap instead.
  setattr(self.__wrapped__, name, value)
/Users/laura/micromamba/envs/devbio-napari-env/lib/python3.9/site-packages/napari/utils/_proxies.py:144: FutureWarning: Labels.color is deprecated since 0.4.19 and will be removed in 0.5.0, please set Labels.colormap directly with an instance of napari.utils.colormaps.DirectLabelColormap instead.
  setattr(self.__wrapped__, name, value)

screenshot = viewer.screenshot(canvas_only=False)

plt.figure(figsize=(15, 10))
plt.imshow(screenshot)
plt.axis('off')
plt.show()

/Users/laura/micromamba/envs/devbio-napari-env/lib/python3.9/site-packages/napari/utils/_proxies.py:144: FutureWarning: Labels.color is deprecated since 0.4.19 and will be removed in 0.5.0, please set Labels.colormap directly with an instance of napari.utils.colormaps.DirectLabelColormap instead.
  setattr(self.__wrapped__, name, value)

# UMAP with all of features
screenshot = viewer.screenshot(canvas_only=False)

plt.figure(figsize=(15, 10))
plt.imshow(screenshot)
plt.axis('off')
plt.show()

# get layer's properties containing UMAP results
df = pd.DataFrame(viewer.layers['tubulin_segmented'].properties)

from scipy.stats import spearmanr

correlations = {}

for column in df.columns:
    
    # don't compute correlation for these columns
    if column in ['label', 'frame', 'UMAP_0', 'UMAP_1', 't-SNE_0', 't-SNE_1', 'index'] or '_CLUSTER_ID' in column:
        continue
        
    corr = spearmanr(df[column], df['UMAP_1'])[0]
    correlations[column] = corr

# sort features based on the absolute value of correlation
sorted_features = sorted(correlations.items(), key=lambda x: abs(x[1]), reverse=True)

for feature, correlation in sorted_features:
    print(f"{feature}: {correlation}")

major_axis_length_tubulin: -0.45427329839716024
eccentricity_tubulin: -0.435280176575601
feret_diameter_max_tubulin: -0.4336607627009672
major_axis_length_actin: -0.4291223605584224
eccentricity_actin: -0.41842276708710285
max_intensity_dapi: 0.4133006019709577
mean_intensity_dapi: 0.40984865036090656
extent_tubulin: 0.4016230410186044
feret_diameter_max_actin: -0.4005739702990752
perimeter_crofton_tubulin: -0.3536617596995489
perimeter_tubulin: -0.352592632881828
extent_actin: 0.3490281141459258
solidity_tubulin: 0.34162244893558275
mean_intensity_tubulin: 0.33198176347249264
perimeter_actin: -0.3278094544133843
perimeter_crofton_actin: -0.3254372727036644
solidity_actin: 0.3024339671147008
eccentricity_dapi: -0.2852678038740644
area_actin: -0.28342198005644476
equivalent_diameter_actin: -0.28342198005644476
max_intensity_tubulin: 0.27599778189154156
min_intensity_dapi: 0.2563386635720779
area_tubulin: -0.24361530735295392
equivalent_diameter_tubulin: -0.24361530735295392
extent_dapi: 0.24306834507608113
feret_diameter_max_dapi: -0.2330346670940936
major_axis_length_dapi: -0.2293713257461209
max_intensity_actin: 0.1871643681205241
mean_intensity_actin: 0.17589117664463438
perimeter_dapi: -0.17016692570674585
perimeter_crofton_dapi: -0.16978422208254318
minor_axis_length_actin: -0.13601508151132968
area_dapi: -0.11274393916665945
equivalent_diameter_dapi: -0.11274393916665945
min_intensity_actin: 0.10658633949816747
orientation_dapi: 0.08256501202213891
solidity_dapi: 0.07849134596027274
min_intensity_tubulin: 0.06977542719412674
minor_axis_length_tubulin: -0.05553353771533256
orientation_actin: 0.03847500917287264
orientation_tubulin: 0.017014170157706402
minor_axis_length_dapi: 0.014483154065609473
Selected labels layer: tubulin_segmented
Selected measurements: ['major_axis_length_tubulin', 'eccentricity_tubulin', 'feret_diameter_max_tubulin', 'major_axis_length_actin', 'max_intensity_dapi', 'mean_intensity_dapi', 'extent_tubulin', 'feret_diameter_max_actin']
Selected clustering method: KMeans
Kmeans_top_corr_featr predictions finished.
Napari status bar display of label properties disabled because https://github.com/napari/napari/issues/5417 and https://github.com/napari/napari/issues/4342

Numba: Attempted to fork from a non-main thread, the TBB library may be in an invalid state in the child process.
/Users/laura/micromamba/envs/devbio-napari-env/lib/python3.9/site-packages/napari/components/viewer_model.py:10: FutureWarning: Labels.color is deprecated since 0.4.19 and will be removed in 0.5.0, please set Labels.colormap directly with an instance of napari.utils.colormaps.DirectLabelColormap instead.
  TYPE_CHECKING,

k-Means Clustering

On most correlating 9 features

screenshot = viewer.screenshot(canvas_only=False)

plt.figure(figsize=(15, 10))
plt.imshow(screenshot)
plt.axis('off')
plt.show()

Plotting these clustering results on the obtained UMAP plot

screenshot = viewer.screenshot(canvas_only=False)

plt.figure(figsize=(15, 10))
plt.imshow(screenshot)
plt.axis('off')
plt.show()

screenshot = viewer.screenshot(canvas_only=False)

plt.figure(figsize=(15, 10))
plt.imshow(screenshot)
plt.axis('off')
plt.show()

Selected labels layer: tubulin_segmented
Selected measurements: ['area_dapi', 'equivalent_diameter_dapi', 'mean_intensity_dapi', 'min_intensity_dapi', 'max_intensity_dapi', 'perimeter_dapi', 'perimeter_crofton_dapi', 'extent_dapi', 'solidity_dapi', 'feret_diameter_max_dapi', 'major_axis_length_dapi', 'minor_axis_length_dapi', 'eccentricity_dapi', 'orientation_dapi', 'area_actin', 'equivalent_diameter_actin', 'mean_intensity_actin', 'min_intensity_actin', 'max_intensity_actin', 'perimeter_actin', 'perimeter_crofton_actin', 'extent_actin', 'solidity_actin', 'feret_diameter_max_actin', 'major_axis_length_actin', 'minor_axis_length_actin', 'eccentricity_actin', 'orientation_actin', 'area_tubulin', 'equivalent_diameter_tubulin', 'mean_intensity_tubulin', 'min_intensity_tubulin', 'max_intensity_tubulin', 'perimeter_tubulin', 'perimeter_crofton_tubulin', 'extent_tubulin', 'solidity_tubulin', 'feret_diameter_max_tubulin', 'major_axis_length_tubulin', 'minor_axis_length_tubulin', 'eccentricity_tubulin', 'orientation_tubulin']
Napari status bar display of label properties disabled because https://github.com/napari/napari/issues/5417 and https://github.com/napari/napari/issues/4342
Dimensionality reduction finished
Selected labels layer: tubulin_segmented
Selected measurements: ['area_dapi', 'equivalent_diameter_dapi', 'mean_intensity_dapi', 'min_intensity_dapi', 'max_intensity_dapi', 'perimeter_dapi', 'perimeter_crofton_dapi', 'extent_dapi', 'solidity_dapi', 'feret_diameter_max_dapi', 'major_axis_length_dapi', 'minor_axis_length_dapi', 'eccentricity_dapi', 'orientation_dapi', 'area_actin', 'equivalent_diameter_actin', 'mean_intensity_actin', 'min_intensity_actin', 'max_intensity_actin', 'perimeter_actin', 'perimeter_crofton_actin', 'extent_actin', 'solidity_actin', 'feret_diameter_max_actin', 'major_axis_length_actin', 'minor_axis_length_actin', 'eccentricity_actin', 'orientation_actin', 'area_tubulin', 'equivalent_diameter_tubulin', 'mean_intensity_tubulin', 'min_intensity_tubulin', 'max_intensity_tubulin', 'perimeter_tubulin', 'perimeter_crofton_tubulin', 'extent_tubulin', 'solidity_tubulin', 'feret_diameter_max_tubulin', 'major_axis_length_tubulin', 'minor_axis_length_tubulin', 'eccentricity_tubulin', 'orientation_tubulin']

/Users/laura/micromamba/envs/devbio-napari-env/lib/python3.9/site-packages/umap/umap_.py:1943: UserWarning: n_jobs value -1 overridden to 1 by setting random_state. Use no seed for parallelism.
  warn(f"n_jobs value {self.n_jobs} overridden to 1 by setting random_state. Use no seed for parallelism.")

Napari status bar display of label properties disabled because https://github.com/napari/napari/issues/5417 and https://github.com/napari/napari/issues/4342
Dimensionality reduction finished

/Users/laura/micromamba/envs/devbio-napari-env/lib/python3.9/site-packages/napari/components/viewer_model.py:10: FutureWarning: Labels.color is deprecated since 0.4.19 and will be removed in 0.5.0, please set Labels.colormap directly with an instance of napari.utils.colormaps.DirectLabelColormap instead.
  TYPE_CHECKING,
/Users/laura/micromamba/envs/devbio-napari-env/lib/python3.9/site-packages/napari/utils/_proxies.py:144: FutureWarning: Labels.color is deprecated since 0.4.19 and will be removed in 0.5.0, please set Labels.colormap directly with an instance of napari.utils.colormaps.DirectLabelColormap instead.
  setattr(self.__wrapped__, name, value)