Examples: Stereo-seq Brain

The test data file locates at https://github.com/bioinfo-biols/RidgeSpace/tree/main/tests/StereoBrain.h5ad.

[1]:

import RidgeSpace
import anndata as ad
import seaborn as sns
sns.set(style="ticks", font_scale=1.5)
adata = ad.read_h5ad('./tests/StereoBrain.h5ad')

Preproccess

Only need run once, including mesh constructing and denosing.

[2]:

RidgeSpace.tl_mesh(adata, obsm_key='X_spatial', mesh_optimization=True, save_address='StereoBrain')

Radius   Count
84.852814     15167
134.164079       88
189.736660       27
216.333077       16
247.386338       15
dtype: int64
Use default radius 86.54987001723342
Triangles mesh saved in ./Ridge_tmp/StereoBrain/
Mesh optimization done

[3]:

RidgeSpace.tl_denoise(adata, plot_name = 'Mbp', use_raw=False, save_address='StereoBrain', iteration=30)
RidgeSpace.tl_denoise(adata, plot_name = 'Adcy5',  use_raw=False, save_address='StereoBrain', iteration=30)
RidgeSpace.tl_denoise(adata, find_obs=True, plot_name = 'vj', save_address='StereoBrain', iteration=15)
RidgeSpace.tl_denoise(adata, plot_name = 'pseudotime', save_address='StereoBrain', iteration=15, find_obs=True)

Denoised value for Mbp saved in ./Ridge_tmp/StereoBrain/
Denoised value for Adcy5 saved in ./Ridge_tmp/StereoBrain/
Denoised value for vj saved in ./Ridge_tmp/StereoBrain/
Denoised value for pseudotime saved in ./Ridge_tmp/StereoBrain/

Plotting

[4]:

RidgeSpace.pl_single(adata, plot_name = 'Mbp', obs_cluster='scc_anno', save_address='StereoBrain', elev=75, view=20, z_height=0.7, lw=0,\
                     bg_alpha=10, plot_clustering=False, normlabels=True, return_ax=True)

[4]:

<Axes3DSubplot:xlabel='Spatial 1', ylabel='Spatial 2'>
_images/StereoBrain_6_1.png 
[5]:

RidgeSpace.pl_multipleIN(adata, plot_nameA = 'Adcy5', plot_nameB = 'Mbp', obs_cluster='scc_anno', save_address='StereoBrain', lw=0, elev=16, view=100,
                         z_height=0.7, height=200, normlabels=True, select_c=['#13e8e1', '#c0e8dc', '#2179b6', '#ff8110'],\
                     plot_clustering=True, clustering_z=[0, 200], clustering_size=13, clustering_alpha=0.7, \
                         bg_alphaA=2, bg_alphaB=2, truncA=10, truncB=10, ylim_min=1000, ylim_max=6200, xlim_max=6800, return_ax=True)

['#13e8e1' '#19c0d1' '#2179b6' '#29a06a' '#8e584d' '#9ae18c' '#ac42fe'
 '#b0c9ea' '#b4c5da' '#b7bf63' '#c0e8dc' '#d8292a' '#e579c4' '#e7cd85'
 '#eb832f' '#ff8110' '#ff9a98' '#ffbd7a']

[5]:

<Axes3DSubplot:xlabel='Spatial 1', ylabel='Spatial 2'>
_images/StereoBrain_7_2.png 
[6]:

RidgeSpace.pl_trajectory(adata, plot_name = 'vj', obs_cluster='scc_anno', save_address='StereoBrain', elev=70, view=20, z_height=0.7,\
                     bg_alpha=0, plot_HE=False, return_ax=True, density_arrow=0.45, min_length=0.3)

[6]:

<Axes3DSubplot:xlabel='Spatial 1', ylabel='Spatial 2'>
_images/StereoBrain_8_1.png 
[7]:

RidgeSpace.pl_trajectory(adata, plot_name = 'pseudotime', save_address='StereoBrain', find_obs=True, obs_cluster='scc_anno', elev=70, view=20, z_height=0.7,\
                     bg_alpha=0, plot_HE=False, return_ax=True, density_arrow=0.45, min_length=0.3)

[7]:

<Axes3DSubplot:xlabel='Spatial 1', ylabel='Spatial 2'>
_images/StereoBrain_9_1.png