6: Visium Data Analysis

Demonstrate the usage of spatialcells for deconvoluted Visium datasets (via cell2location)

@author: Guihong Wan and Boshen Yan
@date: Feb 6 2024
@last updated: Feb 6 2024
[1]:

import pandas as pd
import matplotlib.pyplot as plt
import matplotlib as mpl
import anndata as ad
import scanpy as sc

import spatialcells as spc

Read data

[4]:

adata = ad.read_h5ad("../../data/sp_1.h5ad")
adata

[4]:

AnnData object with n_obs × n_vars = 4035 × 10217
    obs: 'in_tissue', 'array_row', 'array_col', 'sample', '_indices', '_scvi_batch', '_scvi_labels'
    var: 'feature_types', 'genome', 'SYMBOL', 'MT_gene'
    uns: '_scvi_manager_uuid', '_scvi_uuid', 'mod', 'spatial'
    obsm: 'MT', 'means_cell_abundance_w_sf', 'q05_cell_abundance_w_sf', 'q95_cell_abundance_w_sf', 'spatial', 'stds_cell_abundance_w_sf'

We use 5% quantile of the posterior distribution, representing the value of cell abundance that the model has high confidence in (aka ‘at least this amount is present’).

https://cell2location.readthedocs.io/en/latest/notebooks/cell2location_tutorial.html#Visualising-cell-abundance-in-spatial-coordinates

[5]:

adata.obs[adata.uns['mod']['factor_names']] = adata.obsm['q05_cell_abundance_w_sf']

[6]:

adata.obs.columns

[6]:

Index(['in_tissue', 'array_row', 'array_col', 'sample', '_indices',
       '_scvi_batch', '_scvi_labels', 'B_Cycling', 'B_GC_DZ', 'B_GC_LZ',
       'B_GC_prePB', 'B_IFN', 'B_activated', 'B_mem', 'B_naive', 'B_plasma',
       'B_preGC', 'DC_CCR7+', 'DC_cDC1', 'DC_cDC2', 'DC_pDC', 'Endo', 'FDC',
       'ILC', 'Macrophages_M1', 'Macrophages_M2', 'Mast', 'Monocytes', 'NK',
       'NKT', 'T_CD4+', 'T_CD4+_TfH', 'T_CD4+_TfH_GC', 'T_CD4+_naive',
       'T_CD8+_CD161+', 'T_CD8+_cytotoxic', 'T_CD8+_naive', 'T_TIM3+', 'T_TfR',
       'T_Treg', 'VSMC'],
      dtype='object')

[7]:

# find centroids of each spot, scaled according to image size in pixels to enable overlay
# tissue_hires_scalef: A scaling factor that converts pixel positions in the original, full-resolution image
# to pixel positions in tissue_hires_image.png.
scalef = adata.uns["spatial"]["V1_Human_Lymph_Node"]["scalefactors"]["tissue_hires_scalef"]
adata.obs["X_centroid"] = adata.obsm["spatial"][:, 0] * scalef
adata.obs["Y_centroid"] = adata.obsm["spatial"][:, 1] * scalef

[8]:


with mpl.rc_context({'axes.facecolor':  'black',
                     'figure.figsize': [4.5, 5]}):

    sc.pl.spatial(adata, cmap='magma',
                  # show first 8 cell types
                  color=['B_naive',  'T_CD4+_naive', 'T_CD8+_naive',
                         'B_GC_DZ', 'B_GC_LZ', 'B_Cycling', 'T_CD4+_TfH_GC', 'FDC',
                         'B_preGC',],
                  ncols=3, size=1.3,
                  img_key='hires',
                  # limit color scale at 99.2% quantile of cell abundance
                  vmin=0, vmax='p99.2'
                 )
../_images/tutorial_6_tutorial_visium_cell2location_8_0.png 
[10]:

## identify celltype-dense spots.
## dense spots are defined here as spots of more than 90% quantile in cell type abundance
adata.obs[
    adata.uns['mod']['factor_names'] + "_dense"
    ] = adata.obs[
        adata.uns['mod']['factor_names']
        ] > adata.obs[
            adata.uns['mod']['factor_names']
            ].quantile(0.90)

# Here, we are interested in visualizing the spatial arrangement of germinal centers
# We located germinal centers based on germinal center specific B cell subtypes
# Which are identified by the Cell2location tutorial pipeline by Kleshchevnikov et al
adata.obs["COI_community"] = (
    adata.obs["B_Cycling_dense"] |
    adata.obs["B_GC_LZ_dense"] |
    adata.obs["B_GC_DZ_dense"]
).astype(int)

[11]:

# plotting spots dense in B cell subtypes of interest

adata1 = adata.copy()
adata1.obs["B_Cycling_dense"] = adata1.obs["B_Cycling_dense"].astype(int)
adata1.obs["B_GC_LZ_dense"] = adata1.obs["B_GC_LZ_dense"].astype(int)
adata1.obs["B_GC_DZ_dense"] = adata1.obs["B_GC_DZ_dense"].astype(int)


with mpl.rc_context({'axes.facecolor':  'black',
                     'figure.figsize': [4.5, 5]}):

    sc.pl.spatial(adata1, cmap='magma',
                  # show first 8 cell types
                  color=['B_Cycling_dense', 'B_GC_LZ_dense', 'B_GC_DZ_dense', "COI_community"],
                  ncols=2, size=1.3,
                  img_key='hires',
                  # limit color scale at 99.2% quantile of cell abundance
                  vmin=0, vmax='p99.2'
                 )
../_images/tutorial_6_tutorial_visium_cell2location_10_0.png

getBoundary

[12]:

eps_estims = spc.spa.estimateInitialDistance(
    adata, ["COI_community"], sampling_ratio=0.2
)
eps_estims

Computing distances...

[12]:

[33.36114384870609,
 93.11537188982376,
 321.85548625417425,
 1257.6925339706372,
 3761.444077919561]

[13]:

boundary = spc.spa.getBoundary(adata, "COI_community", [1], alpha=eps_estims[0]/2)
# filter out small regions or holes based on the number of edges
# or the area of the region
spot_diameter = adata.uns["spatial"]["V1_Human_Lymph_Node"]["scalefactors"]["spot_diameter_fullres"]
spots_boundary = spc.spa.getExtendedBoundary(boundary, offset=spot_diameter * scalef / 2)

boundary1 = spc.spa.pruneSmallComponents(spots_boundary, min_edges=10, min_area=2000)

fig, [ax1, ax2] = plt.subplots(1, 2, figsize=(10,6))
spc.plt.plotBoundary(boundary, ax=ax1, color="r", label="Centroids Boundary")
ax1.invert_yaxis()
ax1.legend()

spc.plt.plotBoundary(boundary1, ax=ax2, color="b", label="Adjusted Spots Boundary")
ax2.invert_yaxis()
ax2.legend()
plt.show()
../_images/tutorial_6_tutorial_visium_cell2location_13_0.png

assignPointsToRegion

Using one or more generated boundaries, cells can be assigned to different regions for downstream analysis.

[29]:

help(spc.spatial.assignPointsToRegions)

Help on function assignPointsToRegions in module spatialcells.spatial._assignPointsToRegions:

assignPointsToRegions(anndata, boundaries_list, region_names, assigncolumn='region', default='BG')
    Assign points to regions based on the boundaries. The region assignment is
    based on the order of the boundaries, so the innermost region should be the
    first element of boundaries_list.

    :param anndata: Anndata object
    :param boundaries_list: List of boundaries
    :param region_names: List of region names. The order and length should match boundaries_list
    :param assigncolumn: Column name for the region assignment
    :param default: Default region name for points that are not assigned to any region


[30]:

regions = ["Germinal Center"]
boundaries_list = [boundary1]

spc.spatial.assignPointsToRegions(
    adata, boundaries_list, regions, assigncolumn="region", default="Background"
)

4012it [00:01, 2839.96it/s]

Assigned points to region: Germinal Center



[31]:

print("Regions:")
print(adata.obs["region"].cat.categories)
print("\nNumber of points in each region:")
print(adata.obs["region"].value_counts())

Regions:
Index(['Germinal Center', 'Background'], dtype='object')

Number of points in each region:
region
Background         3487
Germinal Center     548
Name: count, dtype: int64

[35]:


# fig, ax = plt.subplots(figsize=(8,10))
# plt.imshow(adata.uns["spatial"]["V1_Human_Lymph_Node"]["images"]["hires"], alpha=1)

# plt.legend(loc="lower left", markerscale=5)
# plt.savefig('HnE.png', dpi=300)

[53]:

adata.obs.columns

[53]:

Index(['in_tissue', 'array_row', 'array_col', 'sample', '_indices',
       '_scvi_batch', '_scvi_labels', 'B_Cycling', 'B_GC_DZ', 'B_GC_LZ',
       'B_GC_prePB', 'B_IFN', 'B_activated', 'B_mem', 'B_naive', 'B_plasma',
       'B_preGC', 'DC_CCR7+', 'DC_cDC1', 'DC_cDC2', 'DC_pDC', 'Endo', 'FDC',
       'ILC', 'Macrophages_M1', 'Macrophages_M2', 'Mast', 'Monocytes', 'NK',
       'NKT', 'T_CD4+', 'T_CD4+_TfH', 'T_CD4+_TfH_GC', 'T_CD4+_naive',
       'T_CD8+_CD161+', 'T_CD8+_cytotoxic', 'T_CD8+_naive', 'T_TIM3+', 'T_TfR',
       'T_Treg', 'VSMC', 'X_centroid', 'Y_centroid', 'B_Cycling_dense',
       'B_GC_DZ_dense', 'B_GC_LZ_dense', 'B_GC_prePB_dense', 'B_IFN_dense',
       'B_activated_dense', 'B_mem_dense', 'B_naive_dense', 'B_plasma_dense',
       'B_preGC_dense', 'DC_CCR7+_dense', 'DC_cDC1_dense', 'DC_cDC2_dense',
       'DC_pDC_dense', 'Endo_dense', 'FDC_dense', 'ILC_dense',
       'Macrophages_M1_dense', 'Macrophages_M2_dense', 'Mast_dense',
       'Monocytes_dense', 'NK_dense', 'NKT_dense', 'T_CD4+_dense',
       'T_CD4+_TfH_dense', 'T_CD4+_TfH_GC_dense', 'T_CD4+_naive_dense',
       'T_CD8+_CD161+_dense', 'T_CD8+_cytotoxic_dense', 'T_CD8+_naive_dense',
       'T_TIM3+_dense', 'T_TfR_dense', 'T_Treg_dense', 'VSMC_dense',
       'COI_community', 'region'],
      dtype='object')

[111]:

# plotting spots dense in B cell subtypes of interest

adata1 = adata.copy()
adata1.obs["B_Cycling_dense"] = adata1.obs["B_Cycling_dense"].astype(int)
adata1.obs["B_GC_LZ_dense"] = adata1.obs["B_GC_LZ_dense"].astype(int)
adata1.obs["B_GC_DZ_dense"] = adata1.obs["B_GC_DZ_dense"].astype(int)

colors = ["orange", "yellow", "green", "tab:blue"]

point_size = 4
fig, ax = plt.subplots(figsize=(8,10))
plt.imshow(adata.uns["spatial"]["V1_Human_Lymph_Node"]["images"]["hires"], alpha=1)

#All
tmp = adata
ax.scatter(*zip(*tmp.obs[["X_centroid", "Y_centroid"]].to_numpy()),
           s=point_size,
           alpha=0.5,
           color=colors[3]
           )

#GC
tmp = adata[adata.obs.B_Cycling_dense|adata.obs.B_GC_LZ_dense|adata.obs.B_Cycling_dense]
ax.scatter(*zip(*tmp.obs[["X_centroid", "Y_centroid"]].to_numpy()),
           s=point_size,
           alpha=0.5,
           color=colors[0]
           )
#B native
tmp = adata[adata.obs.B_naive_dense]
ax.scatter(*zip(*tmp.obs[["X_centroid", "Y_centroid"]].to_numpy()),
           s=point_size,
           alpha=0.5,
           color=colors[1]
           )
#T native
tmp = adata[adata.obs["T_CD4+_naive_dense"]|adata.obs["T_CD8+_naive_dense"]]
ax.scatter(*zip(*tmp.obs[["X_centroid", "Y_centroid"]].to_numpy()),
           s=point_size,
           alpha=0.5,
           color=colors[2]
           )


plt.legend(loc="lower left", markerscale=5)
plt.show()
# plt.savefig('cells.png', dpi=300)

No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
../_images/tutorial_6_tutorial_visium_cell2location_20_1.png 
[110]:

colors = ["tab:orange", "tab:blue", "tab:green", "tab:grey", "tab:red"]

point_size = 4
fig, ax = plt.subplots(figsize=(8,10))
spc.plt.plotBoundary(
    boundary1, color="r", label="Germinal Center Boundary", ax=ax, alpha=0.9
)
# get points in each region in order
for i, region in enumerate(adata.obs["region"].cat.categories):
    tmp = adata[adata.obs.region == region]
    ax.scatter(
        *zip(*tmp.obs[["X_centroid", "Y_centroid"]].to_numpy()),
        s=point_size,
        alpha=0.5,
        label=region,
        color=colors[i]
    )

plt.imshow(adata.uns["spatial"]["V1_Human_Lymph_Node"]["images"]["hires"], alpha=1)

# plt.legend(loc="lower left", markerscale=5)
plt.show()
# plt.savefig('GC.png', dpi=300)
../_images/tutorial_6_tutorial_visium_cell2location_21_0.png 
[85]:

## get overall boundary
overall_boundary = spc.spa.getBoundary(adata, "in_tissue", [1], alpha=20)
# filter out small regions or holes based on the number of edges
# or the area of the region
overall_boundary = spc.spa.getExtendedBoundary(overall_boundary, offset=spot_diameter * scalef / 2)
overall_boundary1 = spc.spa.pruneSmallComponents(overall_boundary, min_edges=10, min_area=1000)

# fig, ax1 = plt.subplots(figsize=(5,6))
# spc.plt.plotBoundary(overall_boundary1, ax=ax1, color="r", label="Overall Boundary")
# ax1.legend()
# ax1.invert_yaxis()
# plt.show()

[174]:

## Helper function to get cell type counts in each region, across several columns
def getRegionCountsMean(adata, regioncol, columns):
    all_counts = []
    for col in columns:
        counts = adata.obs.groupby(regioncol, observed=True)[col].mean()
        all_counts.append(counts)
    return pd.concat(all_counts, axis=1)

def getRegionCounts(adata, regioncol, columns):
    all_counts = []
    for col in columns:
        counts = adata.obs.groupby(regioncol, observed=True)[col].sum()
        all_counts.append(counts)
    return pd.concat(all_counts, axis=1)

[176]:

raw_estim_mean = getRegionCountsMean(adata, "region", adata.uns['mod']['factor_names'])
percentage_dense = getRegionCounts(adata, "region", adata.uns['mod']['factor_names'] + "_dense")

[177]:

adata.uns['mod']['factor_names']

[177]:

array(['B_Cycling', 'B_GC_DZ', 'B_GC_LZ', 'B_GC_prePB', 'B_IFN',
       'B_activated', 'B_mem', 'B_naive', 'B_plasma', 'B_preGC',
       'DC_CCR7+', 'DC_cDC1', 'DC_cDC2', 'DC_pDC', 'Endo', 'FDC', 'ILC',
       'Macrophages_M1', 'Macrophages_M2', 'Mast', 'Monocytes', 'NK',
       'NKT', 'T_CD4+', 'T_CD4+_TfH', 'T_CD4+_TfH_GC', 'T_CD4+_naive',
       'T_CD8+_CD161+', 'T_CD8+_cytotoxic', 'T_CD8+_naive', 'T_TIM3+',
       'T_TfR', 'T_Treg', 'VSMC'], dtype=object)

[178]:

# mean estimated cell type abundance in each spot for each region
raw_estim_mean

[178]:
B_Cycling B_GC_DZ B_GC_LZ B_GC_prePB B_IFN B_activated B_mem B_naive B_plasma B_preGC ... T_CD4+_TfH T_CD4+_TfH_GC T_CD4+_naive T_CD8+_CD161+ T_CD8+_cytotoxic T_CD8+_naive T_TIM3+ T_TfR T_Treg VSMC
region
Germinal Center 3.489793 1.543892 2.812592 0.851064 0.574457 0.831627 1.312993 1.608901 0.718583 0.706312 ... 0.249425 1.695331 0.686543 0.198562 0.316849 0.391157 0.945408 0.515738 0.293817 0.403211
Background 0.665491 0.384414 0.509018 0.240387 0.718301 1.187099 1.957374 1.868405 1.379708 1.140211 ... 0.538775 0.620881 2.324264 0.471576 0.488169 1.266793 0.943437 0.662367 0.734309 0.756567

2 rows × 34 columns

[179]:

# percentage of spots in each region that are
# considered dense with each cell type
percentage_dense

[179]:
B_Cycling_dense B_GC_DZ_dense B_GC_LZ_dense B_GC_prePB_dense B_IFN_dense B_activated_dense B_mem_dense B_naive_dense B_plasma_dense B_preGC_dense ... T_CD4+_TfH_dense T_CD4+_TfH_GC_dense T_CD4+_naive_dense T_CD8+_CD161+_dense T_CD8+_cytotoxic_dense T_CD8+_naive_dense T_TIM3+_dense T_TfR_dense T_Treg_dense VSMC_dense
region
Germinal Center 345 354 365 340 59 53 36 68 1 31 ... 3 269 5 1 5 4 36 5 3 5
Background 59 50 39 64 345 351 368 336 403 373 ... 401 135 399 403 399 400 368 399 401 399

2 rows × 34 columns

[180]:

## areas (in image pixels):
print("Area of germinal centers:", spc.msmt.getRegionArea(boundary1))
print("Area of tissue sample:", spc.msmt.getRegionArea(overall_boundary1))
print(
    "Percentage area of germinal centers: "
    f"{(spc.msmt.getRegionArea(boundary1) / spc.msmt.getRegionArea(overall_boundary1) * 100):.2f} %"
)

Area of germinal centers: 237973.08060695932
Area of tissue sample: 1907465.6849017586
Percentage area of germinal centers: 12.48 %

[182]:

## cell type compositions
# for demo for now, only showing naive b cells
print(spc.msmt.getRegionComposition(adata, "B_naive_dense", regions=["Background"]))
print(spc.msmt.getRegionComposition(adata, "B_naive_dense", regions=["Germinal Center"]))

   B_naive_dense  cell_count  composition
0          False        3151     0.903642
1           True         336     0.096358
   B_naive_dense  cell_count  composition
0          False         480     0.875912
1           True          68     0.124088

[183]:

spc.msmt.getRegionComposition(
    adata, "FDC_dense",
    regions=["Background"],
    regioncol="region"
)

[183]:
FDC_dense cell_count composition
0 False 3411 0.978205
1 True 76 0.021795 
[184]:

spc.msmt.getRegionComposition(
    adata, "FDC_dense",
    regions=["Germinal Center"],
    regioncol="region"
)

[184]:
FDC_dense cell_count composition
0 True 328 0.59854
1 False 220 0.40146 
[185]:

adata.obs.region.value_counts()

[185]:

region
Background         3487
Germinal Center     548
Name: count, dtype: int64

[186]:

boundary_component_list = spc.spa.getComponents(boundary1)
print("Number of components:", len(boundary_component_list))
fig, ax = plt.subplots(figsize=(8, 10))

for i, region in enumerate(adata.obs["region"].cat.categories):
    tmp = adata[adata.obs.region == region]
    ax.scatter(
        *zip(*tmp.obs[["X_centroid", "Y_centroid"]].to_numpy()),
        s=point_size,
        alpha=0.6,
        label=region,
        color=colors[i]
    )

spc.plt.plotRegions(
    boundary_component_list,
    colors_list=["r", "b", "g", "y", "m", "k"],
    x_offset=1,
    y_offset=1,
    ax=ax,
)

ax.imshow(adata.uns["spatial"]["V1_Human_Lymph_Node"]["images"]["hires"], alpha=1)

# ax.invert_yaxis()
plt.show()

Number of components: 33
../_images/tutorial_6_tutorial_visium_cell2location_34_1.png 
[187]:

colors = ["orange", "tab:blue", "green", "tab:blue"]

# draw only the 3 germinal centers highlighted in the original paper
selected_center_boundaries = [
    boundary_component_list[18],
    boundary_component_list[21],
    boundary_component_list[17]
]

fig, ax = plt.subplots(figsize=(8, 8))

for i, region in enumerate(adata.obs["region"].cat.categories):
    tmp = adata[adata.obs.region == region]
    ax.scatter(
        *zip(*tmp.obs[["X_centroid", "Y_centroid"]].to_numpy()),
        s=spot_diameter * scalef * 4,
        alpha=0.6,
        label=region,
        color=colors[i]
    )

spc.plt.plotRegions(
    selected_center_boundaries,
    colors_list=["r", "r", "r", "y", "m", "k"],
    x_offset=1,
    y_offset=1,
    ax=ax,
    linewidth=3,
)

ax.imshow(adata.uns["spatial"]["V1_Human_Lymph_Node"]["images"]["hires"], alpha=1)

ax.set_ylim(600, 50)
ax.set_xlim(750, 1150)
# ax.legend()
# ax.invert_yaxis()
plt.show()
../_images/tutorial_6_tutorial_visium_cell2location_35_0.png 
[189]:

from shapely import MultiPolygon
example_centers_boundary = MultiPolygon([
    boundary_component_list[17].geoms[0],
    boundary_component_list[18].geoms[0],
    boundary_component_list[21].geoms[0],
])

extended_boundary = spc.spa.getExtendedBoundary(example_centers_boundary, offset=spot_diameter * scalef * 1.7)

extended_boundary2 = spc.spa.getExtendedBoundary(extended_boundary, offset=spot_diameter * scalef * 2.5)


regions1 = ["Germinal Center", "Extended Boundary", "Extended Boundary 2"]
boundaries_list1 = [example_centers_boundary, extended_boundary, extended_boundary2]

spc.spatial.assignPointsToRegions(
    adata, boundaries_list1, regions1, assigncolumn="selectRegion", default="Background"
)


fig, ax = plt.subplots(figsize=(8, 10))

# colors = [ "tab:red","tab:blue", "tab:blue", "tab:blue"]
colors = [ "tab:red","yellow", "tab:green", "tab:blue"]

for i, region in enumerate(adata.obs["selectRegion"].cat.categories):
    tmp = adata[adata.obs["selectRegion"] == region]
    ax.scatter(
        *zip(*tmp.obs[["X_centroid", "Y_centroid"]].to_numpy()),
        s=spot_diameter * scalef * 4,
        alpha=0.6,
        label=region,
        color=colors[i]
    )

spc.plt.plotBoundary(
    example_centers_boundary, color="r", label="Germinal Center Boundary", ax=ax, alpha=0.9, linewidth=3
)
spc.plt.plotBoundary(
    extended_boundary, color="yellow", label="Extended Boundary", ax=ax, alpha=0.9, linewidth=3
)
spc.plt.plotBoundary(
    extended_boundary2, color="g", label="Extended Boundary 2", ax=ax, alpha=0.9, linewidth=3
)
ax.invert_yaxis()
# ax.legend()
ax.imshow(adata.uns["spatial"]["V1_Human_Lymph_Node"]["images"]["hires"], alpha=1)
ax.set_ylim(700, 0)
ax.set_xlim(700, 1200)
plt.show()

136it [00:00, 13150.41it/s]

Assigned points to region: Germinal Center

0it [00:00, ?it/s]

114it [00:00, 14200.25it/s]

Assigned points to region: Extended Boundary

160it [00:00, 18058.46it/s]

Assigned points to region: Extended Boundary 2
../_images/tutorial_6_tutorial_visium_cell2location_36_7.png 
[190]:

select_raw_estim_mean = getRegionCounts(adata, "selectRegion", adata.uns['mod']['factor_names'])
select_percentage_dense = getRegionCounts(adata, "selectRegion", adata.uns['mod']['factor_names'] + "_dense")

[191]:

select_raw_estim_mean

[191]:
B_Cycling B_GC_DZ B_GC_LZ B_GC_prePB B_IFN B_activated B_mem B_naive B_plasma B_preGC ... T_CD4+_TfH T_CD4+_TfH_GC T_CD4+_naive T_CD8+_CD161+ T_CD8+_cytotoxic T_CD8+_naive T_TIM3+ T_TfR T_Treg VSMC
selectRegion
Germinal Center 286.902964 113.064673 159.679207 51.245447 23.273735 43.106062 66.509449 79.642546 51.496140 49.612274 ... 17.269178 111.767758 52.921448 12.987905 18.707585 29.394039 61.963411 37.720135 21.032434 25.728872
Extended Boundary 46.554121 27.333019 40.321693 16.738548 40.212077 71.434094 154.702069 140.014732 48.803659 74.105153 ... 26.607056 49.270372 125.810323 21.887633 22.388177 66.206820 52.500484 41.552343 41.759977 34.385885
Extended Boundary 2 87.158340 32.426226 39.669068 24.924477 99.335584 88.471167 144.949332 124.572365 118.408322 152.365655 ... 66.516513 61.398461 371.280319 56.007545 48.345026 195.484460 117.901923 91.279761 101.936597 74.761433
Background 3812.358491 2013.681578 3076.577213 1211.702830 2656.696160 4392.135409 7178.721235 7052.576379 4986.118130 4086.893288 ... 1905.000187 2871.617469 7930.921863 1662.313938 1786.438468 4340.574597 3575.483885 2421.745253 2556.818049 2724.234355

4 rows × 34 columns

[208]:

select_percentage_dense1 = select_percentage_dense.transpose().sort_values(["Germinal Center", "Extended Boundary"], ascending=False).transpose()

[209]:

select_percentage_dense1.columns

[209]:

Index(['B_GC_DZ_dense', 'B_Cycling_dense', 'B_GC_prePB_dense', 'FDC_dense',
       'B_GC_LZ_dense', 'T_CD4+_TfH_GC_dense', 'Macrophages_M1_dense',
       'DC_cDC1_dense', 'B_naive_dense', 'B_activated_dense', 'B_preGC_dense',
       'T_TIM3+_dense', 'B_mem_dense', 'B_IFN_dense', 'T_CD4+_dense',
       'T_TfR_dense', 'T_CD4+_naive_dense', 'T_Treg_dense', 'NKT_dense',
       'NK_dense', 'DC_CCR7+_dense', 'DC_cDC2_dense', 'DC_pDC_dense',
       'Endo_dense', 'T_CD8+_naive_dense', 'T_CD8+_CD161+_dense', 'VSMC_dense',
       'T_CD4+_TfH_dense', 'B_plasma_dense', 'ILC_dense', 'Monocytes_dense',
       'T_CD8+_cytotoxic_dense', 'Macrophages_M2_dense', 'Mast_dense'],
      dtype='object')

[210]:

select_percentage_dense1 =select_percentage_dense[[
    'B_GC_LZ_dense', 'B_GC_DZ_dense', 'B_Cycling_dense', 'B_GC_prePB_dense',
    'FDC_dense', 'T_CD4+_TfH_GC_dense',
    'B_naive_dense', 'B_mem_dense', 'B_activated_dense', 'B_IFN_dense',
    'T_TfR_dense', 'T_CD4+_dense', 'T_CD4+_naive_dense','DC_CCR7+_dense',
    'T_Treg_dense', 'T_CD8+_naive_dense',
    'Monocytes_dense', 'Endo_dense', 'VSMC_dense','Mast_dense', 'NKT_dense', 'B_plasma_dense'
]]

[211]:

# select_percentage_dense1.loc["Germinal Center"] = select_percentage_dense1.loc["Germinal Center"]/select_percentage_dense1.loc["Germinal Center"].sum()

[213]:

fig, axs = plt.subplots(3,1, figsize=(10, 8), sharex=True)

select_percentage_dense1.loc["Germinal Center"].plot(kind="bar", ax=axs[0], color="r")
# axs[0].set_title("Germinal Center", fontsize=12)
axs[0].set_ylim([0, 55])

select_percentage_dense1.loc["Extended Boundary"].plot(kind="bar", ax=axs[1], color="y")
# axs[1].set_title("Extended Boundary Region")
axs[1].set_ylim([0, 55])

select_percentage_dense1.loc["Extended Boundary 2"].plot(kind="bar", ax=axs[2], color="g")
# axs[2].set_title("Extended Boundary 2 Region")
axs[2].set_ylim([0, 55])


# fig.set_ylabel("Percentage of Dense Spots")
# select_percentage_dense1.loc["Background"].plot(kind="bar", ax=axs[3], color="gray")
# axs[3].set_title("Background Region")
# axs[3].set_xlabel("Cell Type")
plt.tight_layout()
plt.show()
../_images/tutorial_6_tutorial_visium_cell2location_43_0.png 
[136]:

# fig, axs = plt.subplots(4,1, figsize=(10, 8), sharex=True)
# select_percentage_dense.loc["Germinal Center"].plot(kind="bar", ax=axs[0], color="r")
# select_percentage_dense.loc["Extended Boundary"].plot(kind="bar", ax=axs[1], color="b")
# select_percentage_dense.loc["Extended Boundary 2"].plot(kind="bar", ax=axs[2], color="g")
# select_percentage_dense.loc["Background"].plot(kind="bar", ax=axs[3], color="k")
# axs[3].set_xlabel("Cell Type")
# plt.show()

[ ]: