Visualization of Nanostring DSP subtissues

The original data of Nanostring DSP technology is different from other technologies, It consists of a number of points to form a ROI, and then the ROI shows the merge expression level. Here we use this tutorial to better visualize the DSP data.

The more complete week 13 slides were used for presentation. Image data can be downloaded from https://nanostring.com/products/geomx-digital-spatial-profiler/spatial-organ-atlas/mouse-development/.

import scanpy as sc
import numpy as np
import pandas as pd
import tifffile as tiff
import matplotlib.pyplot as plt

Read

Read the Anndata data generated in Read spatial omics or Spatiotemporal pattern tutorial.

adata = sc.read_h5ad('/csb2/project/SpatialPackage_whq/Tutorial/mouse_embryo_dsp_adata.h5ad')# write Spatiotemporal pattern tutorial's h5ad to make the 'mouse_embryo_dsp_adata.h5ad' data
image = tiff.imread('/csb2/project/tmp_download/GeoMx_MouseDevelopment/image_files/mu_dev_E13_image_files/mu_dev_E13_006.ome.tiff')
image = image[0, :, :]

adata.obs.columns

Index(['SlideName', 'ScanLabel', 'ROILabel', 'SegmentLabel', 'QCFlags',
       'AOISurfaceArea', 'AOINucleiCount', 'ROICoordinateX', 'ROICoordinateY',
       'RawReads', 'AlignedReads', 'DeduplicatedReads', 'TrimmedReads',
       'StitchedReads', 'SequencingSaturation', 'SequencingSetID', 'UMIQ30',
       'RTSQ30', 'GeoMxNgsPipelineVersion',
       'LOT_Mouse_NGS_Whole_Transcriptome_Atlas_RNA', 'Tissue_annotation',
       'Tissue', 'p63_expression', 'Tissue_substructure', 'Tissue_type',
       'Timepoint', 'ROIID', 'SegmentID', 'ScanWidth', 'ScanHeight',
       'ScanOffsetX', 'ScanOffsetY',
       'LOQ (Mouse NGS Whole Transcriptome Atlas RNA)', 'NormalizationFactor',
       'ExpressionFilteringThreshold (Mouse NGS Whole Transcriptome Atlas RNA)'],
      dtype='object')

adata.uns['point']

	slide	roi	x	y
0	mu_dev_E13_006	1	13011.793535	10484.417086
1	mu_dev_E13_006	1	13109.708338	10499.139178
2	mu_dev_E13_006	1	13184.708338	10547.123928
3	mu_dev_E13_006	1	13242.708338	10642.093747
4	mu_dev_E13_006	1	13261.840745	10730.656697
...	...	...	...	...
4171	mu_dev_E13_011	58	4548.893838	10387.159712
4172	mu_dev_E13_011	58	4477.882657	10362.799483
4173	mu_dev_E13_011	58	4437.660936	10371.299021
4174	mu_dev_E13_011	58	4487.202775	10425.060315
4175	mu_dev_E13_011	58	4636.188087	10423.920083

4176 rows × 4 columns

Preprocessing

We used subtissues of the heart, lung, and midgut for visualization.

adata = adata[np.isin(adata.obs['Tissue'], ['Heart', 'Lung', 'Midgut']), :].copy()
adata.obs['Tissue_s'] = adata.obs['Tissue'].astype(str) + ' ' + adata.obs['Tissue_substructure'].astype(str)

Slide_name = 'mu_dev_E13_006'

tissue_list = ['Heart Trabecula', 'Heart Valve', 'Heart Wall', 'Lung Epithelium',
                'Lung Mesenchyme', 'Midgut Epithelium', 'Midgut Mesenchyme', 'Midgut Neuron']

color_dict = {
    'Heart Trabecula': '#beeac4',
    'Heart Valve': '#800000',
    'Heart Wall': '#fa6d1d',
    'Lung Epithelium': '#4169e1',
    'Lung Mesenchyme': '#ee82ee',
    'Midgut Epithelium': '#da1f18',
    'Midgut Mesenchyme': '#ffe5b4',
    'Midgut Neuron': '#f47a75'
}

# Gray scale transformation
image = image / 23758 * 255*2

obs_inf = adata.obs.copy()
obs_inf = obs_inf[obs_inf['SlideName'] == Slide_name]

point_inf = adata.uns['point'].copy()
point_inf = point_inf[point_inf['slide'] == Slide_name]

points_roi = {}
for index in point_inf['roi'].unique():
    roi = point_inf[point_inf['roi'] == index]
    points_roi[index] = list(zip(roi['x'], roi['y']))

Visualization

Place the original image and the spot distribution of the sub-tissues on two layers and then show them separately.

fig = plt.figure(figsize=(20, 20))

ax = fig.add_axes([0.1, 0.15, 0.7, 0.7])
ax.xaxis.set_ticks_position('top')
ax.invert_yaxis()

# plot subtissues
for row in obs_inf.itertuples():
    subt = getattr(row, 'Tissue_s')
    if subt in tissue_list:
        # print(getattr(row, 'ROILabel'), getattr(row, 'Tissue_substructure'))
        for point in points_roi[getattr(row, 'ROILabel')]:
            x, y = point
            ax.scatter(x, y, c=color_dict[subt], label=subt, s=2, zorder=2)

ax.set_xticks([])
ax.set_yticks([])
ax.set_xlabel('Spatial1', size=20)
ax.set_ylabel('Spatial2', size=20)

handles, labels = ax.get_legend_handles_labels()
unique = [(h, l) for i, (h, l) in enumerate(zip(handles, labels)) if l not in labels[:i]]

ax.legend(*zip(*unique), bbox_to_anchor=(1.01, 0.9), loc='upper left', markerscale=5)

# plot image
ax.imshow(image, zorder=1)

fig.show()

/tmp/ipykernel_74608/2934043088.py:29: UserWarning: FigureCanvasAgg is non-interactive, and thus cannot be shown
  fig.show()