SpatialScope

https://spatialscope-tutorial.readthedocs.io/en/latest/
Admin
SpatialScope is a computational method for inferring single-cell resolution spatial transcriptomics...
Tokens:36,482
Snippets:332
Trust Score:9.6
Update:4 months ago
Show doc for...
Tokens:
Raw
### Check Installation Status

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/installation.html

Commands to verify the installation by checking the help message of the Cell_Type_Identification script.

```bash
$ python ./src/Cell_Type_Identification.py -h
# or directly if `python setup.py develop` has been run
$ Cell_Type_Identification.py -h

```

--------------------------------

### Clone Repository and Set Up Environment

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/installation.html

Commands to clone the SpatialScope repository, create and activate a conda environment, install the package in develop mode, and apply a fix for squidpy.

```bash
$ git clone https://github.com/YangLabHKUST/SpatialScope.git
$ cd SpatialScope
$ conda env create -f environment.yml
$ conda activate SpatialScope
$ python setup.py develop
# fix bug of squidpy
$ rsync ./src/_feature_mixin.py ~/.conda/envs/SpatialScope/lib/python3.9/site-packages/squidpy/im/_feature_mixin.py

```

--------------------------------

### Step 3: Gene expression decomposition

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/usage.html

Command for gene expression decomposition, utilizing a pre-trained model and specifying GPU resources.

```bash
python ./src/Decomposition.py --tissue heart --out_dir  ./output --SC_Data ./Ckpts_scRefs/Heart_D2/Ref_Heart_sanger_D2.h5ad --cell_class_column cell_type  --ckpt_path ./Ckpts_scRefs/Heart_D2/model_5000.pt --spot_range 0,100 --gpu 0,1,2,3

```

--------------------------------

### Step 2: Cell type identification

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/usage.html

Command to identify cell types using segmented nuclei data and a single-cell reference dataset.

```bash
python ./src/Cell_Type_Identification.py --tissue heart --out_dir  ./output  --ST_Data ./output/heart/sp_adata_ns.h5ad --SC_Data ./Ckpts_scRefs/Heart_D2/Ref_Heart_sanger_D2.h5ad --cell_class_column cell_type

```

--------------------------------

### Step 1: Nuclei segmentation

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/usage.html

Command to perform nuclei segmentation. It requires paths to spatial data, image data, and an output directory.

```bash
python ./src/Nuclei_Segmentation.py --tissue heart --out_dir  ./output  --ST_Data ./demo_data/V1_Human_Heart_spatial.h5ad --Img_Data  ./demo_data/V1_Human_Heart_image.tif

```

--------------------------------

### Get MERFISH data shape

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-MOp-MERFISH.html

This code snippet displays the shape of the preprocessed MERFISH data.

```python
ad_sp.shape

```

--------------------------------

### Load Initial Proportions

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Cerebellum-Slideseq.html

Loads initial proportion data from a pickle file.

```python
with open('../output/cere/InitProp.pickle', 'rb') as handle:
    InitProp = pickle.load(handle)
```

--------------------------------

### Cellular Communications - Representative Example 3

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Cerebellum-Slideseq.html

Plots the interaction between 'Nlgn2' and 'Nrxn3' genes in 'Granule' and 'Purkinje' cell types, respectively.

```python
L_gene,R_gene,L_ct,R_ct = 'Nlgn2','Nrxn3','Granule','Purkinje'
fig, ax = plt.subplots(1, 1, figsize=(10.2, 8),dpi=100,facecolor='#fafafa')
PlotLRGenes(ad_scst,L_gene,R_gene,L_ct,R_ct,title='${}$'.format(L_gene+'-'+R_gene),ax=ax,invertY=False,s=20)
```

--------------------------------

### Cellular Communications - Representative Example 2

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Cerebellum-Slideseq.html

Plots the interaction between 'Mdk' and 'Ptprz1' genes in 'Bergmann' and 'Granule' cell types, respectively.

```python
L_gene,R_gene,L_ct,R_ct = 'Mdk','Ptprz1','Bergmann','Granule'
fig, ax = plt.subplots(1, 1, figsize=(10.2, 8),dpi=100,facecolor='#fafafa')
PlotLRGenes(ad_scst,L_gene,R_gene,L_ct,R_ct,title='${}$'.format(L_gene+'-'+R_gene),ax=ax,invertY=False,s=20)
```

--------------------------------

### Cellular Communications - Representative Example 1

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Cerebellum-Slideseq.html

Plots the interaction between 'Apoe' and 'Sorl1' genes in 'MLI2' and 'Purkinje' cell types, respectively.

```python
L_gene,R_gene,L_ct,R_ct = 'Apoe','Sorl1','MLI2','Purkinje'
fig, ax = plt.subplots(1, 1, figsize=(10.2, 8),dpi=100,facecolor='#fafafa')
PlotLRGenes(ad_scst,L_gene,R_gene,L_ct,R_ct,title='${}$'.format(L_gene+'-'+R_gene),ax=ax,invertY=False,s=10)
```

--------------------------------

### Training scRNA-seq Reference Model

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Brain.html

This command initiates the training of a score-based model for scRNA-seq reference data, utilizing multiple GPUs for parallel processing.

```bash
python ./src/Train_scRef.py \
--ckpt_path ./Ckpts_scRefs/VISp \
--scRef ./Ckpts_scRefs/VISp/Ref_scRNA_VISp_qc2_2Kgenes.h5ad \
--cell_class_column cell_subclass \
--gpus 0,1,2,3
```

--------------------------------

### Import necessary libraries

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Brain.html

Imports common Python libraries for data manipulation, plotting, and scientific computing, along with custom utilities.

```python
import numpy as np
import pandas as pd
import pathlib
import matplotlib.pyplot as plt
import matplotlib as mpl
import seaborn as sns
import scanpy as sc

import sys
sys.path.append('../src')
from utils import *

import warnings
warnings.filterwarnings('ignore')
```

--------------------------------

### Import PASTE and related libraries

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Brain.html

Imports the PASTE library for spatial alignment and other necessary scientific computing libraries.

```python
import paste as pst
from scipy.spatial.distance import pdist
import matplotlib.patches as mpatches
```

--------------------------------

### Get statistical significance of gene pair expression changes upon cell-cell interaction

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Human-Heart-CCell-interaction.html

This code snippet calculates the statistical significance of gene pair expression changes considering spatial interactions between cells.

```R
spatial_all_scores = spatCellCellcom(my_giotto_object,
                                     spatial_network_name = 'Delaunay_network',
                                     cluster_column = 'cell_types',
                                     random_iter = 500,
                                     gene_set_1 = select_ligands,
                                     gene_set_2 = select_receptors,
                                     adjust_method = 'fdr',
                                     do_parallel = T,
                                     cores = 4,
                                     verbose = 'none')
```

--------------------------------

### Running SpatialScope Gene Expression Decomposition

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Brain.html

These commands execute the gene expression decomposition step of the SpatialScope pipeline, processing different ranges of spots and utilizing specified GPUs.

```bash
python ./src/Decomposition.py --tissue cortex --out_dir  ./output --SC_Data ./Ckpts_scRefs/VISp/Ref_scRNA_VISp_qc2_2Kgenes.h5ad --cell_class_column cell_subclass  --ckpt_path ./Ckpts_scRefs/VISp/model_5000.pt --spot_range 0,500 --gpu 2,3,4,8

python ./src/Decomposition.py --tissue cortex --out_dir  ./output --SC_Data ./Ckpts_scRefs/VISp/Ref_scRNA_VISp_qc2_2Kgenes.h5ad --cell_class_column cell_subclass  --ckpt_path ./Ckpts_scRefs/VISp/model_5000.pt --spot_range 500,1000 --gpu 2,3,4,8

python ./src/Decomposition.py --tissue cortex --out_dir  ./output --SC_Data ./Ckpts_scRefs/VISp/Ref_scRNA_VISp_qc2_2Kgenes.h5ad --cell_class_column cell_subclass  --ckpt_path ./Ckpts_scRefs/VISp/model_5000.pt --spot_range 1000,1715 --gpu 1,2,4,8
```

--------------------------------

### Loading Data

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Brain.html

Loads two .h5ad files, slice1 and slice2, which have undergone nuclei segmentation.

```python
slice1 = sc.read(slice1_file)
slice2 = sc.read(slice2_file)
```

--------------------------------

### Get statistical significance of gene pair expression changes based on expression

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Human-Heart-CCell-interaction.html

This code snippet calculates the statistical significance of gene pair expression changes based solely on expression levels within cell clusters.

```R
expr_only_scores = exprCellCellcom(gobject = my_giotto_object,
                                   cluster_column = 'cell_types',
                                   random_iter = 500,
                                   gene_set_1 = select_ligands,
                                   gene_set_2 = select_receptors)
```

--------------------------------

### Loading and Preprocessing Slice 2 Data

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Brain.html

Loads the AnnData object for 'slice2' and applies an exponential transformation to its expression matrix.

```python
slice2_file = '../demo_data/Visium_MouceBrain_Cortex_section2.h5ad'
slice2 = sc.read(slice2_file)
slice2.X = np.exp(slice2.X.A)-1
```

--------------------------------

### Initialize ImageContainer

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Human-Heart.html

Loads the human heart image into an ImageContainer for spatial analysis.

```python
import squidpy as sq

image = plt.imread('../demo_data/V1_Human_Heart_image.tif')
img = sq.im.ImageContainer(image)
```

--------------------------------

### Load and preprocess scRNA-seq reference data

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Brain.html

Loads a scRNA-seq reference dataset, filters it for the 'VISp' region and specific cell subclasses, and prepares it for downstream analysis.

```python
ad_sc = sc.read('adata_processed_sc.h5ad')
ad_sc = ad_sc[ad_sc.obs['dissected_region']=='VISp']
ad_sc = ad_sc[ad_sc.obs['cell_subclass']!='nan']
ad_sc = ad_sc[ad_sc.obs['cell_subclass']!='No Class']
ad_sc.obs['cell_subclass'] = ad_sc.obs['cell_subclass'].astype(str)
ad_sc.obs.loc[ad_sc.obs.cell_subclass=='NP','cell_subclass'] = 'L5 NP'

cell_type_column = 'cell_subclass'

ad_sc.X.max(),ad_sc.shape
```

--------------------------------

### Stack slices pairwise

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Brain.html

Stacks slices pairwise after alignment.

```python
slices, pis = [slice1, slice2], [pi12]
new_slices = pst.stack_slices_pairwise(slices, pis)
```

--------------------------------

### Create Giotto Object

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Human-Heart-CCell-interaction.html

Initializes a Giotto object with raw expression data, spatial locations, cell metadata, and Python path for instructions.

```R
my_instructions = createGiottoInstructions(python_path='/home/xwanaf/.conda/envs/rbayes/bin/python')
my_giotto_object = createGiottoObject(raw_exprs = path_to_matrix,
                                      spatial_locs = path_to_locations,
                                      cell_metadata = my_cell_metadata,
                                      instructions = my_instructions)
my_giotto_object@cell_metadata$cell_types <- my_giotto_object@cell_metadata$V1
```

--------------------------------

### Load sampled cells and initialize plotting

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Human-Heart.html

Loads sampled cells using scRef and sets up plotting context.

```python
sampled_cells = sc.read('../Ckpts_scRefs/Heart_D2/model_5000.h5ad') # 2K cells sampled from the learned gene expression distribution of scRef
cell_type_key='cell_type'
sns.set_context('paper',font_scale=1.5)
ada_all = PlotSampledData(sampled_cells,ad_sc,cell_type_key,palette=color_dict)
```

--------------------------------

### Import Libraries

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-MOp-MERFISH.html

Imports necessary Python libraries for data analysis and visualization, including scanpy, numpy, pandas, and matplotlib.

```python
import numpy as np
import pandas as pd
import pathlib
import matplotlib.pyplot as plt
import matplotlib as mpl
import seaborn as sns
import scanpy as sc
import pickle

import sys
sys.path.append('../src')
from utils import *

import warnings
warnings.filterwarnings('ignore')

```

--------------------------------

### Writing AnnData to File

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Brain.html

Saves the processed AnnData object to a HDF5 file.

```python
slice_3D.write('../output/cortex/sp_adata_ns.h5ad')
```

--------------------------------

### Run Cell Type Identification

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Brain.html

This command executes the cell type identification script with specified parameters for the cortex dataset.

```bash
./src/Cell_Type_Identification.py --tissue cortex --out_dir  ./output  --ST_Data ./output/cortex/sp_adata_ns.h5ad --SC_Data ./Ckpts_scRefs/VISp/Ref_scRNA_VISp_qc2_2Kgenes.h5ad --cell_class_column cell_subclass --nu 15

```

--------------------------------

### Import Libraries

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Human-Heart.html

Imports necessary Python libraries for data analysis and visualization, including numpy, pandas, matplotlib, seaborn, and scanpy.

```python
import numpy as np
import pandas as pd
import pathlib
import matplotlib.pyplot as plt
import matplotlib as mpl
import seaborn as sns
import scanpy as sc

import sys
sys.path.append('../src')
from utils import *

import warnings
warnings.filterwarnings('ignore')

```

--------------------------------

### Prepare slices for alignment

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Brain.html

This code snippet prepares two slices, slice1 and slice2, for alignment by modifying their index and cell_locations attributes. It ensures that the names are unique and appends slice identifiers to them.

```python
slice1.obs.index = ['slice1-'+_ for _ in slice1.obs.index]
slice2.obs.index = ['slice2-'+_ for _ in slice2.obs.index]
slice1.uns['cell_locations']['slice_index'] = 'slice1'
slice2.uns['cell_locations']['slice_index'] = 'slice2'
slice1.uns['cell_locations']['spot_index'] = slice1.uns['cell_locations']['spot_index'].map(lambda x:'slice1-'+x)
slice2.uns['cell_locations']['spot_index'] = slice2.uns['cell_locations']['spot_index'].map(lambda x:'slice2-'+x)
slice1.uns['cell_locations']['cell_index'] = slice1.uns['cell_locations']['cell_index'].map(lambda x:'slice1-'+x)
slice2.uns['cell_locations']['cell_index'] = slice2.uns['cell_locations']['cell_index'].map(lambda x:'slice2-'+x)

slice1.obs_names_make_unique()
slice2.obs_names_make_unique()
slice1.var_names_make_unique()
slice2.var_names_make_unique()
```

--------------------------------

### Load Visium Data

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Brain.html

Loads two sections of Visium mouse brain cortex data using scanpy.

```python
slice1_file = '../demo_data/Visium_MouseBrain_Cortex_section1.h5ad'
slice2_file = '../demo_data/Visium_MouseBrain_Cortex_section2.h5ad'
slice1 = sc.read(slice1_file)
slice2 = sc.read(slice2_file)

```

--------------------------------

### Train scRNA-seq reference model

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-MOp-MERFISH.html

This command trains a scRNA-seq reference model using four GPUs. The checkpoints and sampled pseudo-cells will be saved in the specified ckpt_path.

```bash
python ./src/Train_scRef.py \
--ckpt_path ./Ckpts_scRefs/MOp \
--scRef ./Ckpts_scRefs/MOp/Ref_snRNA_mop_qc3_2Kgenes.h5ad \
--cell_class_column subclass_label \
--gpus 0,1,2,3

```

--------------------------------

### Data Preparation for Visualization

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Brain.html

Extracts x, y, z coordinates for cells from two different slices and assigns them to separate variables.

```python
xyz1 = celltype_res.loc[celltype_res['slice_index']=='slice1',['x','y','z']].values
xyz2 = celltype_res.loc[celltype_res['slice_index']=='slice2',['x','y','z']].values

x1,y1,z1 = xyz1[:,0],xyz1[:,1],xyz1[:,2]
x2,y2,z2 = xyz2[:,0],xyz2[:,1],xyz2[:,2]
```

--------------------------------

### Load Libraries

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Human-Heart-CCell-interaction.html

Loads the necessary Giotto and data.table libraries.

```R
library(Giotto)
library(data.table)
```

--------------------------------

### Load spatial transcriptomics slices

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Brain.html

Loads two slices of Visium Mouse Brain Cortex data for alignment.

```python
slice1_file = '../demo_data/Visium_MouseBrain_Cortex_section1.h5ad'
slice2_file = '../demo_data/Visium_MouseBrain_Cortex_section2.h5ad'
```

--------------------------------

### Log transform and perform PCA

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Brain.html

Applies log transformation to the data and performs Principal Component Analysis (PCA).

```python
sc.pp.log1p(ad_sc)
sc.pp.pca(ad_sc,svd_solver='arpack', n_comps=30, use_highly_variable=True)
ad_sc.X.max()
```

--------------------------------

### Plot Visium Cells for Slice 2

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Brain.html

Generates a visualization of cells in slice2, colored by their identified cell type.

```python
fig, ax = plt.subplots(1,1,figsize=(10, 8),dpi=130)
PlotVisiumCells(slice2,"discrete_label_ct",size=0.4,alpha_img=0.1,lw=0.7,palette=color_dict,ax=ax)

```

--------------------------------

### Compute neighbors and UMAP

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Brain.html

Calculates the neighborhood graph and performs Uniform Manifold Approximation and Projection (UMAP) for visualization.

```python
sc.pp.neighbors(ad_sc, metric='cosine', n_neighbors=30, n_pcs = 30)
sc.tl.umap(ad_sc, min_dist = 0.3, spread = 1, maxiter=100)
```

--------------------------------

### Import Libraries

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Cerebellum-Slideseq.html

Imports necessary Python libraries for data manipulation, plotting, and single-cell analysis.

```python
import numpy as np
import pandas as pd
import pathlib
import matplotlib.pyplot as plt
import matplotlib as mpl
import seaborn as sns
import scanpy as sc

import sys
sys.path.append('../src')
from utils import *
import pickle

import warnings
warnings.filterwarnings('ignore')
```

--------------------------------

### Load and Preprocess scRNA-seq Reference

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Human-Heart.html

Loads the raw scRNA-seq data, filters cells and genes based on quality metrics, selects a specific donor, and samples cells to balance cell type representation.

```python
ad_sc = sc.read('global_raw.h5ad')

ad_sc = ad_sc[ad_sc.obs['cell_type']!='doublets']
ad_sc = ad_sc[ad_sc.obs['cell_type']!='NotAssigned']
ad_sc = ad_sc[ad_sc.obs['cell_type']!='Mesothelial']
ad_sc = ad_sc[ad_sc.obs['cell_source']=='Sanger-Nuclei']

cell_type_column = 'cell_type'

sc.pp.filter_cells(ad_sc, min_counts=500)
sc.pp.filter_cells(ad_sc, max_counts=20000)
sc.pp.filter_genes(ad_sc, min_cells=5)
ad_sc = ad_sc[:,~np.array([_.startswith('MT-') for _ in ad_sc.var.index])]
ad_sc = ad_sc[:,~np.array([_.startswith('mt-') for _ in ad_sc.var.index])]

ad_sc = ad_sc[ad_sc.obs['donor']=='D2'].copy() # reduce batch effect among doners
ad_sc = ad_sc[ad_sc.obs.index.isin(ad_sc.obs.groupby('cell_type').apply(
    lambda x: x.sample(frac=3000/x.shape[0],replace=False) if x.shape[0]>3000 else x).reset_index(level=0,drop=True).index)].copy()

ad_sc.X.max(),ad_sc.shape

```

--------------------------------

### Plot aligned slices

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Brain.html

Plots the aligned slices with specified colors and legend.

```python
slice_colors = ['#e41a1c','#377eb8']
slice_colors = ['teal', 'orange']
slice_colors = ['#377eb8', 'orange']
plt.figure(figsize=(11,7))
for i in range(len(new_slices)):
    pst.plot_slice(new_slices[i],slice_colors[i],s=350)
plt.legend(handles=[mpatches.Patch(color=slice_colors[0], label='Cortex slice 1'),mpatches.Patch(color=slice_colors[1], label='Cortex slice 2')],
           fontsize=29,bbox_to_anchor=(.96,.72))
plt.gca().invert_yaxis()
plt.axis('off')
plt.show()
```

--------------------------------

### Write AnnData to File

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Human-Heart.html

Saves the processed AnnData object to a .h5ad file.

```python
ad_scst_crop.write('../demo_data/human_heart_ROI_generated_cells.h5ad')
```

--------------------------------

### Train scRNA-seq reference model

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Human-Heart.html

Command to train a scRNA-seq reference model using four GPUs. It specifies checkpoint path, reference data, cell class column, and training parameters like sigma values and learning rate step.

```python
python ./src/Train_scRef.py \
--ckpt_path ./Ckpts_scRefs/Heart_D2 \
--scRef ./Ckpts_scRefs/Heart_D2/Ref_Heart_sanger_D2.h5ad \
--cell_class_column cell_type \
--gpus 0,1,2,3 \
--sigma_begin 50 --sigma_end 0.002 --step_lr 3e-7

```

--------------------------------

### Plot Visium Cells for Slice 1

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Brain.html

Generates a visualization of cells in slice1, colored by their identified cell type.

```python
fig, ax = plt.subplots(1,1,figsize=(10, 8),dpi=130)
PlotVisiumCells(slice1,"discrete_label_ct",size=0.4,alpha_img=0.1,lw=0.7,palette=color_dict,ax=ax)

```

--------------------------------

### Load Human Heart Data

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Human-Heart.html

Loads the human heart dataset from a specified H5AD file.

```python
ad_scst_crop = sc.read('../demo_data/human_heart_ROI_generated_cells.h5ad')
```

--------------------------------

### Define Color Dictionary for Cell Types

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Brain.html

A dictionary mapping cell type names to specific colors for visualization.

```python
color_dict = {'Astro': '#1f77b4',
 'CR': '#aec7e8',
 'Endo': '#ffbb78',
 'L2/3 IT': '#ff7f0e',
 'L4': '#2ca02c',
 'L5 IT': '#bcbd22',
 'L5 NP': '#d62728',
 'L5 PT': '#ff9896',
 'L6 CT': '#17becf',
 'L6 IT': '#c5b0d5',
 'L6b': '#8c564b',
 'Lamp5': '#c49c94',
 'Macrophage': '#e377c2',
 'Meis2': '#f7b6d2',
 'Oligo': '#7f7f7f',
 'Peri': '#c7c7c7',
 'Pvalb': '#98df8a',
 'SMC': '#dbdb8d',
 'Serpinf1': '#9467bd',
 'Sncg': '#9edae5',
 'Sst': '#82A8CE',
 'VLMC': '#B87BCE',
 'Vip': '#3FC241'}

```

--------------------------------

### Plotting Visium Cells for Slice 2

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Brain.html

Generates a plot of Visium cells for slice2, visualizing cell numbers with specified parameters.

```python
fig, ax = plt.subplots(1,1,figsize=(12, 6),dpi=100)
PlotVisiumCells(slice2,"cell_nums",0.23,0.35,0.3,ax=ax)
```

--------------------------------

### Display Cell Locations Head

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Brain.html

Displays the first few rows of the cell locations DataFrame for slice1.

```python
slice1.uns['cell_locations'].head()

```

--------------------------------

### Crop and display image section

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Human-Heart.html

Crops a specific corner of the image and displays it channel-wise.

```python
crop = img.crop_corner(0, 0)
crop.show(channelwise=True,layer='image')
```

--------------------------------

### Incorporate ligand-receptor genes and update markers

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Brain.html

Adds ligand-receptor genes to the list of markers and updates the 'Marker' and 'highly_variable' attributes in the AnnData object.

```python
ligand_recept = list(set(pd.read_csv('../extdata/ligand_receptors.txt',sep='	').melt()['value'].values))
# if scRNA-seq reference is from human tissue, run following code to make gene name consistent
# ligand_recept = [_.upper() for _ in ligand_recept]
markers = markers+ligand_recept

ad_sc.var.loc[ad_sc.var.index.isin(markers),'Marker'] = True
ad_sc.var['Marker'] = ad_sc.var['Marker'].fillna(False)
ad_sc.var['highly_variable'] = ad_sc.var['Marker']
```

--------------------------------

### Original spatial coordinates

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Brain.html

Displays the original spatial coordinates of slice1.

```python
slice1.obsm['spatial']
```

--------------------------------

### Visualize UMAP projection

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Brain.html

Generates a UMAP plot colored by cell subclass to visualize the clustering of cell types.

```python
fig, axs = plt.subplots(1, 1, figsize=(10, 10))
sc.pl.umap(
    ad_sc, color=cell_type_column, size=10, frameon=False, show=False, ax=axs,legend_loc='on data'
)
plt.tight_layout()
```

--------------------------------

### Loading and Concatenating ST Data

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Brain.html

Loads multiple AnnData objects representing different slices of mouse brain ST data and concatenates them into a single object. It then filters for data from 'slice2'.

```python
ad_scsts_list = []
ad_scsts_list.append(sc.read('../output/cortex/generated_cells_spot0_500.h5ad'))
ad_scsts_list.append(sc.read('../output/cortex/generated_cells_spot500_1000.h5ad'))
ad_scsts_list.append(sc.read('../output/cortex/generated_cells_spot1000_1715.h5ad'))

ad_scsts = ad_scsts_list[0].concatenate(
        ad_scsts_list[1:],
        batch_key="_",
        uns_merge="unique",
        index_unique=None
        )

ad_scst_slice2 = ad_scsts[ad_scsts.obs.slice_index=='slice2']
```

--------------------------------

### Plotting Visium Cells for Slice 1

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Brain.html

Generates a plot of Visium cells for slice1, visualizing cell numbers with specified parameters.

```python
fig, ax = plt.subplots(1,1,figsize=(12, 6),dpi=100)
PlotVisiumCells(slice1,"cell_nums",0.23,0.35,0.3,ax=ax)
```

--------------------------------

### Displaying Cell Locations

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Brain.html

Displays the head of the 'cell_locations' DataFrame from slice1, showing spatial coordinates and cell indices.

```python
slice1.uns['cell_locations'].head()
```

--------------------------------

### Plotting Sampled Data from scRef and Pseudo-cells

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Brain.html

This code snippet visualizes sampled pseudo-cells against the scRNA-seq reference data, using a specified cell type key and color palette.

```python
ad_sc = sc.read('../Ckpts_scRefs/VISp/Ref_scRNA_VISp_qc2_2Kgenes.h5ad')
sampled_cells = sc.read('../Ckpts_scRefs/VISp/model_5000.h5ad') # 2K cells sampled from the learned gene expression distribution of scRef
cell_type_key='cell_subclass'
sns.set_context('paper',font_scale=1.5)
adata_all = PlotSampledData(sampled_cells,ad_sc,cell_type_key,palette=color_dict)
```

--------------------------------

### Benchmarking imputation accuracy

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-MOp-MERFISH.html

Shell script to run the decomposition script for benchmarking imputation accuracy across different spot ranges.

```bash
set -ex;
declare -a arr=("0,1000" "1000,2000" "2000,3000" "3000,4000" "4000,5000" "5000,5551")
for i in "${arr[@]}"
doprint
python ./src/Decomposition.py
–tissue merfish
–out_dir ./output
–SC_Data ./Ckpts_scRefs/MOp/Ref_snRNA_mop_qc3_2Kgenes.h5ad
–cell_class_column subclass_label
–ckpt_path ./Ckpts_scRefs/MOp/model_5000.pt
–spot_range $i
–replicates 5
–leave_out_test
–test_genes Cux2,Otof,Rorb,Rspo1,Sulf2,Fezf2,Osr1
–gpu 2,3,4,8
done
```

--------------------------------

### Visualize UMAP by Cell States

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Human-Heart.html

Generates a UMAP visualization colored by 'cell_states' to explore potential sub-populations or states within the cell types.

```python
fig, axs = plt.subplots(1, 1, figsize=(10, 10))
sc.pl.umap(
    ad_sc, color="cell_states", size=10, frameon=False, show=False, ax=axs,legend_loc='on data'
)
plt.tight_layout()

```

--------------------------------

### Data Transformation and Visualization

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-MOp-MERFISH.html

Applies exponential transformation if necessary and visualizes the distribution of total counts per cell.

```python
if ad_sc.X.max()<20:
    ad_sc.X = np.exp(ad_sc.X)-1
plt.hist(ad_sc.X.sum(1),bins=100)
plt.show()

```

--------------------------------

### Identify highly variable and marker genes

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Brain.html

Identifies highly variable genes and cell type marker genes, combining them for further analysis.

```python
sc.pp.highly_variable_genes(ad_sc, flavor='seurat_v3',n_top_genes=1000)
ad_sc.uns['log1p'] = {}
ad_sc.uns['log1p']['base'] = None
sc.tl.rank_genes_groups(ad_sc, groupby=cell_type_column, method='wilcoxon')
markers_df = pd.DataFrame(ad_sc.uns["rank_genes_groups"]["names"]).iloc[0:30, :]
markers = list(np.unique(markers_df.melt().value.values))
markers = list(set(ad_sc.var.loc[ad_sc.var['highly_variable']==1].index)|set(markers))  # highly variable genes + cell type marker genes
len(markers)
```

--------------------------------

### Display Cell Type Results Head

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Brain.html

Displays the first few rows of the cell type results DataFrame.

```python
celltype_res.head()

```

--------------------------------

### Plotting Visium Cells

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Brain.html

This code snippet visualizes Visium cells from a spatial transcriptomics slice, subsetting by a specific cell class and using a defined color palette.

```python
fig, ax = plt.subplots(1,1,figsize=(10, 8),dpi=130)
l_class = ['L4']#, 'L6 CT', 'L6b']
PlotVisiumCells(slice2,"discrete_label_ct",size=0.7,alpha_img=0.1,lw=0.7,subset=l_class,palette=color_dict,ax=ax)
```

--------------------------------

### Load and process ligand-receptor data

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Human-Heart-CCell-interaction.html

This code snippet loads mouse ligand-receptor data, processes it to identify detected ligands and receptors in the dataset, and selects them for further analysis.

```R
LR_data = data.table::fread(system.file("extdata", "mouse_ligand_receptors.txt", package = 'Giotto'))

LR_data$L_upper = toupper(LR_data$mouseLigand)
LR_data$R_upper = toupper(LR_data$mouseReceptor)
LR_data[, ligand_det := ifelse(L_upper %in% my_giotto_object@gene_ID, T, F)]
LR_data[, receptor_det := ifelse(R_upper %in% my_giotto_object@gene_ID, T, F)]
LR_data_det = LR_data[ligand_det == T & receptor_det == T]
select_ligands = LR_data_det$L_upper
select_receptors = LR_data_det$R_upper
```

--------------------------------

### Gene expression decomposition (first 1000 spots)

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Human-Heart.html

Command to perform gene expression decomposition for the first 1000 spots, specifying tissue, output directory, single-cell data, checkpoint path, spot range, and GPUs.

```python
python ./src/Decomposition.py --tissue heart --out_dir  ./output --SC_Data ./Ckpts_scRefs/Heart_D2/Ref_Heart_sanger_D2.h5ad --cell_class_column cell_type  --ckpt_path ./Ckpts_scRefs/Heart_D2/model_5000.pt --spot_range 0,1000 --gpu 0,1,2,3

```

--------------------------------

### Comparing Max Expression Values

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Brain.html

Compares the maximum expression values between the preprocessed 'slice2' data and the filtered 'ad_scst_slice2' data.

```python
slice2.X.max(),ad_scst_slice2.X.max()
```

--------------------------------

### Gene expression decomposition (spots 1000-2000)

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Human-Heart.html

Command to perform gene expression decomposition for spots 1000 to 2000.

```python
python ./src/Decomposition.py --tissue heart --out_dir  ./output --SC_Data ./Ckpts_scRefs/Heart_D2/Ref_Heart_sanger_D2.h5ad --cell_class_column cell_type  --ckpt_path ./Ckpts_scRefs/Heart_D2/model_5000.pt --spot_range 1000,2000 --gpu 0,1,2,3

```

--------------------------------

### Aligned spatial coordinates

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Brain.html

Displays the aligned spatial coordinates of slice1.

```python
slice1.obsm['spatial_align']
```

--------------------------------

### Load AnnData object

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-MOp-MERFISH.html

Loads the spatial data from a specified HDF5 file into an AnnData object.

```python
ad_sp = sc.read('../output/merfish/sp_adata.h5ad')
```

--------------------------------

### Plotting Visium Cells (L6 CT subset)

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Brain.html

Similar to the previous snippet, this visualizes Visium cells but subsets for the 'L6 CT' cell class.

```python
fig, ax = plt.subplots(1,1,figsize=(10, 8),dpi=130)
l_class = ['L6 CT']#, 'L6 CT', 'L6b']
PlotVisiumCells(slice2,"discrete_label_ct",size=0.7,alpha_img=0.1,lw=0.7,subset=l_class,palette=color_dict,ax=ax)
```

--------------------------------

### Combine aligned cell locations

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Brain.html

Combines the aligned cell locations from slice1 and slice2.

```python
spatial_align = np.vstack((slice1.obsm['spatial_align'],slice2.obsm['spatial_align']))
cell_locations_align = slice1.uns['cell_locations'].append(slice2.uns['cell_locations']).reset_index(drop=True)
```

--------------------------------

### Prepare data for cell-cell interaction analysis

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Human-Heart.html

Saves the expression matrix, coordinates, and cell types to files for downstream cell-cell interaction analysis.

```python
ad_scst_crop = sc.read('../demo_data/human_heart_ROI_generated_cells.h5ad')
ad_scst_crop.to_df().T.to_csv('../output/heart/express.txt', sep='	')
ad_scst_crop.obs.loc[:,['x', 'y']].to_csv('../output/heart/coords.txt', sep='	', index = False)
ad_scst_crop.obs['discrete_label_ct'].to_csv('../output/heart/cell_type.txt', sep='	', index = False, header = False)
```

--------------------------------

### Preprocessing MERFISH data

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-MOp-MERFISH.html

This code snippet reads MERFISH data, renames columns, creates cell location information, and saves the preprocessed data.

```python
ad_sp = sc.read('../demo_data/MERFISH_mop.h5ad')

ad_sp.obs = ad_sp.obs.rename(columns = {'X':'x', 'Y':'y'})
cell_locations = ad_sp.obs.copy()
cell_locations['spot_index'] = np.array(cell_locations.index)
cell_locations['cell_index'] = cell_locations['spot_index'].map(lambda x:x+'_0')
cell_locations['cell_nums'] = np.ones(cell_locations.shape[0]).astype(int)
ad_sp.uns['cell_locations'] = cell_locations

ad_sp.write('../demo_data/MERFISH_mop_preprocess.h5ad')

```

--------------------------------

### Concatenate MERFISH .h5ad files

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-MOp-MERFISH.html

Concatenates multiple MERFISH .h5ad files into a single AnnData object.

```python
ad_res = ConcatCells(s=0,e=6000,inter=1000,es=5551,file_path='../output/merfish/',prefix='generated_cells_spot',suffix='.h5ad')
```

--------------------------------

### Pairwise align the slices

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Brain.html

Aligns two slices pairwise using the `pairwise_align` function.

```python
pi12 = pst.pairwise_align(slice1, slice2)
```

--------------------------------

### Accessing Spatial Data

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Brain.html

Retrieves the spatial alignment data stored in the 'spatial' attribute of the .obsm accessor.

```python
slice_3D.obsm['spatial']
```

--------------------------------

### Train Score-Based Model

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Cerebellum-Slideseq.html

Trains a score-based model using four GPUs, specifying checkpoint path, scRNA-seq reference data, cell class column, and GPU IDs.

```bash
python ./src/Train_scRef.py \
--ckpt_path ./Ckpts_scRefs/Cerebellum \
--scRef ./Ckpts_scRefs/Cerebellum/Ref_snRNA_cerebellum_qc.h5ad \
--cell_class_column cell_type \
--gpus 0,1,2,3
```

--------------------------------

### Load Cell Type Results

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Brain.html

Reads the cell type identification results from a CSV file into a pandas DataFrame.

```python
celltype_res = pd.read_csv('../output/cortex/CellTypeLabel_nu15.csv',index_col=0)

```

--------------------------------

### Load and preprocess single-cell data

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Human-Heart.html

Loads a dataset, filters for Smooth Muscle Cells, logs the data, and prepares it for analysis. It also loads a reference dataset, filters it, and renames cell states.

```python
ad_scst_crop = sc.read('../demo_data/human_heart_ROI_generated_cells.h5ad')

smscells_gx = ad_scst_crop[ad_scst_crop.obs['discrete_label_ct']=='Smooth_muscle_cells'].copy()
smscells_gx.obs['cell_type'] = 'Generated_SMC_Cells'
smscells_gx.obs['cell_states'] = 'Generated_SMC_Cells'
sc.pp.log1p(smscells_gx)

ad_sc = sc.read('../Ckpts_scRefs/Heart_D2/Ref_Heart_sanger_D2.h5ad')
ad_sc = ad_sc[ad_sc.obs['cell_type'].isin(['Endothelial','Smooth_muscle_cells'])]
ad_sc = ad_sc[~ad_sc.obs['cell_states'].isin(['EC4_immune','EC10_CMC-like','EC9_FB-like','EC8_ln'])]
ad_sc = ad_sc[:,ad_sc.var.Marker]
rename_dict = {
    'EC1_cap':'EC_cap',
    'EC3_cap':'EC_cap',
    'EC2_cap':'EC_cap'
}

ad_sc.obs['cell_states'] = ad_sc.obs['cell_states'].map(lambda x:rename_dict[x] if x in rename_dict.keys() else x)
```

--------------------------------

### Visualize UMAP results

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Human-Heart.html

Generates a UMAP plot colored by cell states to visualize the clustering of different cell types.

```python
fig, axs = plt.subplots(1, 1, figsize=(11, 10),dpi=100)
plt.rcParams["legend.markerscale"] = 1.3
sns.set_context('paper',font_scale=2.2)
sc.pl.umap(
    adata_all, color="cell_states", size=80, frameon=False, palette='Set1', alpha=0.7, show=False, ax=axs#,legend_loc='on data'
)
plt.tight_layout()
```

--------------------------------

### Inspect AnnData Object

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Human-Heart.html

Displays the structure and summary of the AnnData object, including observations, variables, and unstructured data.

```python
ad_scst_crop
```

--------------------------------

### KDE Plot for Right Gene Expression (Slide-seq vs SpatialScope)

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Cerebellum-Slideseq.html

Generates a Kernel Density Estimate plot for the expression of the right gene in a specific cell type, comparing Slide-seq and SpatialScope data.

```python
gene = R_gene
celltype = R_ct
sns.set_context('paper',font_scale=2.4)
fig, ax = plt.subplots(1, 1, figsize=(8, 6),dpi=100)
sns.kdeplot(data=ad_sp[ad_sp.obs['discrete_label_ct']==celltype,gene].to_df(), x=gene,ax=ax,fill=True,color='#D6D657',label='Slide-seq')
sns.kdeplot(data=ad_scst[ad_scst.obs['discrete_label_ct']==celltype,gene].to_df(), x=gene,ax=ax,fill=True,color='#BF8AD0',label='SpatialScope',clip=[-1,10])
ax.set_xlabel('${}$ expression in {}'.format(gene,celltype))
#plt.legend(loc='best')
ax.spines['top'].set_visible(False)
ax.spines['right'].set_visible(False)
plt.show()
```

--------------------------------

### Assign aligned spatial coordinates

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Brain.html

Assigns the aligned spatial coordinates to the slices.

```python
slice1.obsm['spatial_align'] = new_slices[0].obsm['spatial']
slice2.obsm['spatial_align'] = new_slices[1].obsm['spatial']
```

--------------------------------

### PCA and Neighbor Graph Construction

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Human-Heart.html

Applies log transformation, performs Principal Component Analysis (PCA) using highly variable genes, and constructs a neighbor graph for dimensionality reduction.

```python
ad_sc.var.loc[ad_sc.var.index.isin(markers),'Marker'] = True
ad_sc.var['Marker'] = ad_sc.var['Marker'].fillna(False)
ad_sc.var['highly_variable'] = ad_sc.var['Marker']

sc.pp.log1p(ad_sc)
sc.pp.pca(ad_sc,svd_solver='arpack', n_comps=30, use_highly_variable=True)

ad_sc.X.max()

```

```python
sc.pp.neighbors(ad_sc, metric='cosine', n_neighbors=30, n_pcs = 30)
sc.tl.umap(ad_sc, min_dist = 0.5, spread = 1, maxiter=60)

```

--------------------------------

### Gene expression decomposition (spots 2000-3000)

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Human-Heart.html

Command to perform gene expression decomposition for spots 2000 to 3000.

```python
python ./src/Decomposition.py --tissue heart --out_dir  ./output --SC_Data ./Ckpts_scRefs/Heart_D2/Ref_Heart_sanger_D2.h5ad --cell_class_column cell_type  --ckpt_path ./Ckpts_scRefs/Heart_D2/model_5000.pt --spot_range 2000,3000 --gpu 0,1,2,3

```

--------------------------------

### UMAP Visualization

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-MOp-MERFISH.html

Generates a UMAP plot colored by cell type to visualize the clustering of cell populations.

```python
fig, axs = plt.subplots(1, 1, figsize=(10, 10))
sc.pl.umap(
    ad_sc, color=cell_type_column, size=15, frameon=False, show=False, ax=axs,legend_loc='on data'
)
plt.tight_layout()

```

--------------------------------

### 3D Scatter Plot of Cell Types

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Brain.html

Generates a 3D scatter plot to visualize cell locations and types for two slices, using a predefined color map.

```python
from matplotlib import pyplot as plt
from mpl_toolkits.mplot3d import Axes3D


fig = plt.figure(figsize=(10,5),dpi=200)
ax = fig.gca(projection='3d')# fig.add_subplot(projection='3d')


ax.get_proj = lambda: np.dot(Axes3D.get_proj(ax), np.diag([0.85, 1.6, 0.8, 1]))

ax.scatter(x1, y1, z1, c=slice1.uns['cell_locations']['discrete_label_ct'].map(color_dict), s=1.5, alpha=1)
ax.scatter(x2, y2, z2, c=slice2.uns['cell_locations']['discrete_label_ct'].map(color_dict), s=1.5, alpha=1)

ax.set_xlabel('$x$')
ax.set_ylabel('$y$')
ax.set_zlabel('$z$')

#ax.set_box_aspect([np.ptp(i) for i in data])  # equal aspect ratio
```

--------------------------------

### Accessing Maximum Value in AnnData X

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Brain.html

Retrieves the maximum value from the .X attribute of the AnnData object.

```python
slice_3D.X.max()
```

--------------------------------

### Update spatial coordinates for alignment

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-Brain.html

Updates the spatial coordinates for alignment by adding a new dimension.

```python
new_slices[0].obsm['spatial'] = np.hstack((new_slices[0].obsm['spatial'],
                                          np.zeros((new_slices[0].obsm['spatial'].shape[0],1))))
new_slices[1].obsm['spatial'] = np.hstack((new_slices[1].obsm['spatial'],
                                          (pdist(new_slices[0].obsm['spatial']).min()**2/2)**.5/2 * np.ones((new_slices[1].obsm['spatial'].shape[0],1))))
```

--------------------------------

### Color Dictionary and Label Names

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Mouse-MOp-MERFISH.html

Defines a color dictionary for mapping cell types to specific colors and a list of label names for cell types.

```python
color_dict = {'Astro': '#1f77b4',
 'Endo': '#aec7e8',
 'L2/3 IT': '#ff7f0e',
 'L5 ET': '#ffbb78',
 'L5 IT': '#2ca02c',
 'L5/6 NP': '#98df8a',
 'L6 CT': '#d62728',
 'L6 IT': '#ff9896',
 'L6b': '#9467bd',
 'Lamp5': '#c5b0d5',
 'Micro': '#8c564b',
 'Oligo': '#c49c94',
 'OPC': '#e377c2',
 'Peri': '#f7b6d2',
 'Pvalb': '#7f7f7f',
 'PVM': '#a3a2a2',
 'Sncg': '#bcbd22',
 'Sst': '#dbdb8d',
 'Vip': '#17becf',
 'VLMC': '#9edae5',
 'None': '#dbd9d9'}

labelnames = ['Astro', 'Endo', 'L2/3 IT', 'L5 IT', 'L5 ET', 'L5/6 NP', 'L6 IT', 'L6 CT', 'L6b', 'Lamp5', 'Micro', 'Oligo', 'OPC', 'Peri', 'Pvalb',
'PVM', 'Sncg', 'Sst', 'Vip', 'VLMC']
```

--------------------------------

### Display cropped image object

Source: https://spatialscope-tutorial.readthedocs.io/en/latest/notebooks/Human-Heart.html

Displays the ImageContainer object after cropping.

```python
crop
```
Tokens:
Raw