Dissecting cell types for anatomical structures - Lung Epithelial Cells (advanced)

This how-to walks through the process of extracting cell types for anatomical structures. It illustrates some of the challenges and nuances of ontology extraction in domains such as anatomy.

In theory it should be as simple as selecting a structure (e.g lung or the respiratory airway) and then selecting the cell types that are part of that structure. In practice, the way that ontologies are constructed renders results unsatisfactory:

The results are often incomplete because more ubiquitous cell types may not be axiomatized as part of a structure.
The results contain many redundancies because ontologies don’t populate things at consistent “levels”

For this example, we’ll work through lung epithelial cells.

An example of a biologist-friendly structure we might want to extract:

Initial exploration and motivation

As a motivating example, let’s look at goblet cell (CL:0000160) in CL. For now we’re going to use the command line, but later in this guide we’ll explore the python interface.

We’ll look at is-a and part-of ancestors of this cell type

[1]:

!runoak -i sqlite:obo:cl viz -p i,p "goblet cell" -o output/goblet-cell.png

What might be surprising to people not familiar with ontologies is the generic placement.

Initial setup

We’ll require some core OAK libraries, as well as pandas

[2]:

from oaklib import get_adapter
from oaklib.query import subclass_of, descendant_of, non_redundant, gap_fill
from oaklib.query import ancestor_of
from oaklib.datamodels.vocabulary import HAS_PART, PART_OF, IS_A

[3]:

import pandas as pd

Connect to the Cell Ontology

We’ll use the sqlite version of CL:

[4]:

cl = get_adapter("sqlite:obo:cl")

[4]:

We will also define some helper functions to convert queries to sets and dataframes

[5]:

def q2set(q):
    """execute query and return as set."""
    return set(q.execute(cl, labels=True))

def q2df(q):
    """execute query and return as dataframe."""
    return pd.DataFrame(q.execute(cl, labels=True))

Setting up queries

Here we will make use of query objects to represent the queries we want to execute. This allows us to build up complex queries in a more readable way.

We will start by defining a query for epithelial cells

[6]:

epithelial_q = subclass_of("epithelial cell")

We can look at the results:

[7]:

q2df(epithelial_q)

[7]:

	0	1
0	CL:0000059	ameloblast
1	CL:0000060	odontoblast
2	CL:0000065	ependymal cell
3	CL:0000066	epithelial cell
4	CL:0000067	ciliated epithelial cell
...	...	...
722	CL:4033037	mucus secreting cell of tracheobronchial tree ...
723	CL:4033044	deuterosomal cell
724	CL:4033048	respiratory suprabasal cell
725	CL:4033049	taste receptor cell of tongue
726	CL:4040003	fetal pre-type II pneumocyte

727 rows × 2 columns

There are a lot of epithelial cells in the ontology, at different levels of cell type specificity, as well as different levels of anatomical specificity.

We’re interested in lung epithelial cells, so let’s define a query to get airway parts - we can intersect this with the epithelial cell query to get lung epithelial cells later:

[8]:

airway_q = descendant_of("respiratory airway", predicates=[PART_OF])
q2df(airway_q)

[8]:

	0	1
0	CL:0000082	epithelial cell of lung
1	CL:0000158	club cell
2	CL:0000307	tracheal epithelial cell
3	CL:0000322	pneumocyte
4	CL:0000583	alveolar macrophage
...	...	...
270	UBERON:8600002	mucosa of lobular bronchiole
271	UBERON:8600003	smooth muscle tissue of lobular bronchiole
272	UBERON:8600010	bronchial submucosal gland ciliated duct
273	UBERON:8600011	tracheal submucosal gland ciliated duct
274	UBERON:8600017	bronchopulmonary segment

275 rows × 2 columns

[9]:

## Boolean queries - AND

Our result includes immune cells and many cell types, so next we will intersect out queries to get lung epithelial cells.

Note we could do this simply by using python to intersect our two existing results, but we will instead do this by making an AND (intersection) query. This is a good practice to get into, as it will allow us to build up more complex queries later.

[10]:

aec_q = epithelial_q & airway_q
aec = q2df(aec_q)
aec

[10]:

	0	1
0	CL:0000082	epithelial cell of lung
1	CL:0000158	club cell
2	CL:0000307	tracheal epithelial cell
3	CL:0000322	pneumocyte
4	CL:0000853	olfactory epithelial supporting cell
...	...	...
61	CL:4033023	airway submucosal gland collecting duct epithe...
62	CL:4033037	mucus secreting cell of tracheobronchial tree ...
63	CL:4033044	deuterosomal cell
64	CL:4033048	respiratory suprabasal cell
65	CL:4040003	fetal pre-type II pneumocyte

66 rows × 2 columns

Non-redundant sets

Here we have fewer rows, but our results still have a lot of redundancy. For example, we have both type I and II pneumocytes, as well as the grouping class pneumocyte.

We will filter the results for redundancy (i.e leaf nodes). As we will see later this is not always the best strategy, as sometimes leaf nodes can be too specific, but it’s a good start for exploration

[11]:

nr_aec_q = non_redundant(aec_q, predicates=[IS_A])
nr_aec = q2df(nr_aec_q)
nr_aec

[11]:

	0	1
0	CL:0000158	club cell
1	CL:0000853	olfactory epithelial supporting cell
2	CL:0002062	type I pneumocyte
3	CL:0002063	type II pneumocyte
4	CL:0002171	globose cell of olfactory epithelium
5	CL:0002184	basal proper cell of olfactory epithelium
6	CL:0002205	brush cell of lobular bronchiole
7	CL:0002206	brush cell of terminal bronchiole
8	CL:0002207	brush cell of trachea
9	CL:0002209	intermediate epitheliocyte
10	CL:0002330	undifferentiated cell of bronchus epithelium
11	CL:0002332	ciliated cell of the bronchus
12	CL:0002480	nasal mucosa goblet cell
13	CL:0002638	bronchioalveolar stem cell
14	CL:0008026	open tracheal system tracheocyte
15	CL:0017000	pulmonary ionocyte
16	CL:1000329	tracheal goblet cell
17	CL:1000348	basal cell of epithelium of trachea
18	CL:1000349	basal cell of epithelium of bronchus
19	CL:1000352	basal cell of epithelium of lobular bronchiole
20	CL:1000377	dense-core granulated cell of epithelium of tr...
21	CL:1000492	mesothelial cell of parietal pleura
22	CL:1000493	mesothelial cell of visceral pleura
23	CL:1001573	nasopharyngeal epithelial cell
24	CL:2000094	nasal cavity respiratory epithelium epithelial...
25	CL:4028002	alveolar capillary type 1 endothelial cell
26	CL:4028003	alveolar capillary type 2 endothelial cell
27	CL:4030023	respiratory hillock cell
28	CL:4033003	myoepithelial cell of bronchus submucosal gland
29	CL:4033007	brush cell of epithelium of lobar bronchus
30	CL:4033009	goblet cell of epithelium of lobar bronchus
31	CL:4033010	neuroendocrine cell of epithelium of lobar bro...
32	CL:4033020	mucus secreting cell of trachea gland
33	CL:4033021	myoepithelial cell of trachea gland
34	CL:4033022	mucus secreting cell of bronchus submucosal gland
35	CL:4033023	airway submucosal gland collecting duct epithe...
36	CL:4033044	deuterosomal cell
37	CL:4033048	respiratory suprabasal cell
38	CL:4040003	fetal pre-type II pneumocyte

[12]:

lung_grouping_subset = ["pneumocyte", "lower respiratory tract", "bronchiole", "respiratory airway", "bronchus", "UBERON:0003126", "alveolus of lung", "upper respiratory tract"]
nr_aec_subset_q = gap_fill(nr_aec_q.execute(cl) + lung_grouping_subset, predicates=[IS_A, PART_OF])
graph = cl.relationships_to_graph(nr_aec_subset_q.execute(cl))

[13]:

from oaklib.utilities.obograph_utils import graph_to_tree_display

print(graph_to_tree_display(graph))

* [] UBERON:0001005 ! respiratory airway
    * [p] UBERON:0001558 ! lower respiratory tract
        * [p] UBERON:0003126 ! trachea
            * [p] CL:4033021 ! myoepithelial cell of trachea gland
            * [p] CL:4033020 ! mucus secreting cell of trachea gland
            * [p] CL:1000377 ! dense-core granulated cell of epithelium of trachea
            * [p] CL:1000348 ! basal cell of epithelium of trachea
            * [p] CL:1000329 ! tracheal goblet cell
            * [p] CL:0002207 ! brush cell of trachea
        * [p] UBERON:0002299 ! alveolus of lung
            * [p] CL:4028003 ! alveolar capillary type 2 endothelial cell
            * [p] CL:4028002 ! alveolar capillary type 1 endothelial cell
            * [p] CL:0000322 ! pneumocyte
                * [i] CL:4040003 ! fetal pre-type II pneumocyte
                * [i] CL:0002063 ! type II pneumocyte
                * [i] CL:0002062 ! type I pneumocyte
        * [p] UBERON:0002186 ! bronchiole
            * [p] CL:1000352 ! basal cell of epithelium of lobular bronchiole
            * [p] CL:0002638 ! bronchioalveolar stem cell
            * [p] CL:0002206 ! brush cell of terminal bronchiole
            * [p] CL:0002205 ! brush cell of lobular bronchiole
        * [p] UBERON:0002185 ! bronchus
            * [p] CL:4033022 ! mucus secreting cell of bronchus submucosal gland
            * [p] CL:4033010 ! neuroendocrine cell of epithelium of lobar bronchus
            * [p] CL:4033009 ! goblet cell of epithelium of lobar bronchus
            * [p] CL:4033007 ! brush cell of epithelium of lobar bronchus
            * [p] CL:4033003 ! myoepithelial cell of bronchus submucosal gland
            * [p] CL:1000349 ! basal cell of epithelium of bronchus
            * [p] CL:0002332 ! ciliated cell of the bronchus
            * [p] CL:0002330 ! undifferentiated cell of bronchus epithelium
        * [p] CL:1000493 ! mesothelial cell of visceral pleura
        * [p] CL:1000492 ! mesothelial cell of parietal pleura
        * [p] CL:0017000 ! pulmonary ionocyte
        * [p] CL:0002209 ! intermediate epitheliocyte
        * [p] CL:0000158 ! club cell
    * [p] UBERON:0001557 ! upper respiratory tract
        * [p] CL:2000094 ! nasal cavity respiratory epithelium epithelial cell of viscerocranial mucosa
        * [p] CL:1001573 ! nasopharyngeal epithelial cell
        * [p] CL:0002480 ! nasal mucosa goblet cell
        * [p] CL:0002184 ! basal proper cell of olfactory epithelium
        * [p] CL:0002171 ! globose cell of olfactory epithelium
        * [p] CL:0000853 ! olfactory epithelial supporting cell
    * [p] CL:4033048 ! respiratory suprabasal cell
    * [p] CL:4033044 ! deuterosomal cell
    * [p] CL:4033023 ! airway submucosal gland collecting duct epithelial cell
    * [p] CL:4030023 ! respiratory hillock cell
    * [p] CL:0008026 ! open tracheal system tracheocyte

Dissecting anatomical structures

TODO: complete this section. It may be the case that the current structure of Uberon is not best placed to demonstrate dissection queries for the respiratory system.

[14]:

dissection_q = ancestor_of(airway_q, predicates=[HAS_PART]) & epithelial_q
dissection = q2df(dissection_q)
dissection

[14]:

	0	1
0	CL:0000066	epithelial cell
1	CL:0000067	ciliated epithelial cell
2	CL:0000071	blood vessel endothelial cell
3	CL:0000075	columnar/cuboidal epithelial cell
4	CL:0000076	squamous epithelial cell
...	...	...
82	CL:4033023	airway submucosal gland collecting duct epithe...
83	CL:4033037	mucus secreting cell of tracheobronchial tree ...
84	CL:4033044	deuterosomal cell
85	CL:4033048	respiratory suprabasal cell
86	CL:4040003	fetal pre-type II pneumocyte

87 rows × 2 columns

[15]:

from matplotlib_venn import venn2
import matplotlib.pyplot as plt

[16]:

venn2([q2set(dissection_q), q2set(aec_q)], set_labels=["Dissection", "AEC"])

[16]:

<matplotlib_venn._common.VennDiagram at 0x7fa1f1306fa0>

../../../_images/examples_Ontologies_CL_Lung-Epithelial-Cell-Types_28_1.png

[17]:

plt.figure(figsize=(24, 16))
venn_axis = venn2([q2set(dissection_q), q2set(aec_q)], set_labels=["Dissection", "AEC"])
only_set1 = q2set(dissection_q) - q2set(aec_q)
only_set2 = q2set(aec_q) - q2set(dissection_q)
common_set = q2set(dissection_q) & q2set(aec_q)
venn_axis.get_label_by_id('10').set_text('\n'.join(map(str, only_set1)))
venn_axis.get_label_by_id('01').set_text('\n'.join(map(str, only_set2)))
venn_axis.get_label_by_id('11').set_text('\n'.join(map(str, common_set)))

# Adjust position and size of the labels (optional)
for id in ('10', '01', '11'):
    label = venn_axis.get_label_by_id(id)
    if label:
        label.set_fontsize(8)

../../../_images/examples_Ontologies_CL_Lung-Epithelial-Cell-Types_29_0.png

[33]:

def dissection_venn(struct: str):
    plt.figure(figsize=(24, 16))
    my_dissection_q = ancestor_of(descendant_of(struct, predicates=[PART_OF]), predicates=[HAS_PART]) & epithelial_q
    venn_axis = venn2([q2set(my_dissection_q), q2set(aec_q)], set_labels=["Dissection", "AEC"])
    only_set1 = q2set(my_dissection_q) - q2set(aec_q)
    only_set2 = q2set(aec_q) - q2set(my_dissection_q)
    common_set = q2set(my_dissection_q) & q2set(aec_q)
    venn_axis.get_label_by_id('10').set_text('\n'.join(map(str, only_set1)))
    venn_axis.get_label_by_id('01').set_text('\n'.join(map(str, only_set2)))
    venn_axis.get_label_by_id('11').set_text('\n'.join(map(str, common_set)))

    # Adjust position and size of the labels (optional)
    for id in ('10', '01', '11'):
        label = venn_axis.get_label_by_id(id)
        if label:
            label.set_fontsize(8)

dissection_venn("bronchus")

../../../_images/examples_Ontologies_CL_Lung-Epithelial-Cell-Types_30_0.png

[34]:

dissection_venn("trachea")

../../../_images/examples_Ontologies_CL_Lung-Epithelial-Cell-Types_31_0.png

Roll ups

[18]:

from oaklib.utilities.axioms.logical_definition_utilities import roll_up_to_genus

aec_minimal = {roll_up_to_genus(t, cl, [PART_OF]) for t in aec_q.execute(cl)}
aec_minimal_df = pd.DataFrame((t, cl.label(t)) for t in aec_minimal)
aec_minimal_df

[18]:

	0	1
0	CL:0002062	type I pneumocyte
1	CL:4033044	deuterosomal cell
2	CL:0000630	supporting cell
3	CL:0000064	ciliated cell
4	CL:0002145	ciliated columnar cell of tracheobronchial tree
5	CL:1000441	epithelial cell of viscerocranial mucosa
6	CL:0002638	bronchioalveolar stem cell
7	CL:0002139	endothelial cell of vascular tree
8	CL:0002184	basal proper cell of olfactory epithelium
9	CL:4028003	alveolar capillary type 2 endothelial cell
10	CL:0002329	basal epithelial cell of tracheobronchial tree
11	CL:0002167	olfactory epithelial cell
12	CL:0002076	endo-epithelial cell
13	CL:0002066	Feyrter cell
14	CL:0000076	squamous epithelial cell
15	CL:4030024	hillock cell
16	CL:0017000	pulmonary ionocyte
17	CL:4040003	fetal pre-type II pneumocyte
18	CL:0002171	globose cell of olfactory epithelium
19	CL:0000160	goblet cell
20	CL:4033023	airway submucosal gland collecting duct epithe...
21	CL:4033048	respiratory suprabasal cell
22	CL:0000066	epithelial cell
23	CL:0000165	neuroendocrine cell
24	CL:0000322	pneumocyte
25	CL:0002633	respiratory basal cell
26	CL:0002330	undifferentiated cell of bronchus epithelium
27	CL:0002209	intermediate epitheliocyte
28	CL:0002063	type II pneumocyte
29	CL:0000319	mucus secreting cell
30	CL:0002204	brush cell
31	CL:4033037	mucus secreting cell of tracheobronchial tree ...
32	CL:4028002	alveolar capillary type 1 endothelial cell
33	CL:0000158	club cell
34	CL:0000077	mesothelial cell
35	CL:0000185	myoepithelial cell

[22]:

LOCS = ["lower respiratory tract", "upper respiratory tract", "trachea", "bronchus", "alveolus", "lung", "nose"]

by_loc = {loc: (subclass_of("epithelial cell") & descendant_of(loc, predicates=[PART_OF])).execute(cl) for loc in LOCS}
print(len(by_loc))
by_genus = {cell: subclass_of(cell).execute(cl) for cell in aec_minimal}
print(len(by_genus))

7
36

---------------------------------------------------------------------------
AttributeError                            Traceback (most recent call last)
Cell In[22], line 11
      9     row = {}
     10     for loc in LOCS:
---> 11         row[loc] = len(by_loc[loc].intersection(by_genus[cell]))
     12     rows.append(row)
     14 xp_df = pd.DataFrame(rows)

AttributeError: 'list' object has no attribute 'intersection'

[28]:

rows = []
for cell in aec_minimal:
    row = {"id": cell, "label": cl.label(cell)}
    for loc in LOCS:
        num = len(set(by_loc[loc]).intersection(by_genus[cell]))
        row[loc] = 1 if num > 0 else 0
    rows.append(row)

xp_df = pd.DataFrame(rows)
xp_df

[28]:

	id	label	lower respiratory tract	upper respiratory tract	trachea	bronchus	alveolus	lung	nose
0	CL:0002062	type I pneumocyte	1	0	0	0	1	1	0
1	CL:4033044	deuterosomal cell	0	0	0	0	0	0	0
2	CL:0000630	supporting cell	0	1	0	0	0	0	1
3	CL:0000064	ciliated cell	1	0	0	1	0	0	0
4	CL:0002145	ciliated columnar cell of tracheobronchial tree	1	0	0	1	0	0	0
5	CL:1000441	epithelial cell of viscerocranial mucosa	0	1	0	0	0	0	1
6	CL:0002638	bronchioalveolar stem cell	1	0	0	0	0	1	0
7	CL:0002139	endothelial cell of vascular tree	1	0	0	0	1	1	0
8	CL:0002184	basal proper cell of olfactory epithelium	0	1	0	0	0	0	1
9	CL:4028003	alveolar capillary type 2 endothelial cell	1	0	0	0	1	1	0
10	CL:0002329	basal epithelial cell of tracheobronchial tree	1	0	0	1	0	0	0
11	CL:0002167	olfactory epithelial cell	0	1	0	0	0	0	1
12	CL:0002076	endo-epithelial cell	1	1	1	1	0	1	0
13	CL:0002066	Feyrter cell	1	0	1	0	0	0	0
14	CL:0000076	squamous epithelial cell	1	0	0	0	1	1	0
15	CL:4030024	hillock cell	0	0	0	0	0	0	0
16	CL:0017000	pulmonary ionocyte	1	0	0	0	0	1	0
17	CL:4040003	fetal pre-type II pneumocyte	1	0	0	0	1	1	0
18	CL:0002171	globose cell of olfactory epithelium	0	1	0	0	0	0	1
19	CL:0000160	goblet cell	1	1	1	1	0	1	1
20	CL:4033023	airway submucosal gland collecting duct epithe...	0	0	0	0	0	0	0
21	CL:4033048	respiratory suprabasal cell	0	0	0	0	0	0	0
22	CL:0000066	epithelial cell	1	1	1	1	1	1	1
23	CL:0000165	neuroendocrine cell	1	0	1	1	0	1	0
24	CL:0000322	pneumocyte	1	0	0	0	1	1	0
25	CL:0002633	respiratory basal cell	1	0	1	1	0	1	0
26	CL:0002330	undifferentiated cell of bronchus epithelium	1	0	0	1	0	0	0
27	CL:0002209	intermediate epitheliocyte	1	0	0	0	0	0	0
28	CL:0002063	type II pneumocyte	1	0	0	0	1	1	0
29	CL:0000319	mucus secreting cell	1	1	1	1	0	1	1
30	CL:0002204	brush cell	1	0	1	1	0	1	0
31	CL:4033037	mucus secreting cell of tracheobronchial tree ...	1	0	1	1	0	0	0
32	CL:4028002	alveolar capillary type 1 endothelial cell	1	0	0	0	1	1	0
33	CL:0000158	club cell	1	0	0	0	0	0	0
34	CL:0000077	mesothelial cell	1	0	0	0	0	1	0
35	CL:0000185	myoepithelial cell	1	0	1	1	1	0	0

[29]:

import pandas as pd
import seaborn as sns
import matplotlib.pyplot as plt


# Assuming df is your dataframe with 'id' and 'label' columns followed by columns of anatomical entities
# Set 'id' or 'label' as the index if it makes it easier to visualize
xp_df.set_index('label', inplace=True)

plt.figure(figsize=(10, 8))
sns.heatmap(xp_df.iloc[:, 1:], cmap='viridis', cbar=True, linewidths=.5)
plt.xlabel('Anatomical Entities')
plt.ylabel('Cell Types')
plt.title('Heatmap of Cell Types across Anatomical Entities')
plt.show()

../../../_images/examples_Ontologies_CL_Lung-Epithelial-Cell-Types_36_0.png

[46]:

hp_by_loc = {}
for loc in LOCS:
    hp_by_loc[loc] = set(ancestor_of(descendant_of(loc, predicates=[PART_OF]) - subclass_of("cell"), predicates=[HAS_PART]).execute(cl))

[47]:

rows = []
for cell in aec_minimal:
    row = {"id": cell, "label": cl.label(cell)}
    cells = set(subclass_of(cell).execute(cl))
    for loc in LOCS:
        num = len(cells.intersection(hp_by_loc[loc]))
        row[loc] = 1 if num > 0 else 0
    rows.append(row)


hp_df = pd.DataFrame(rows)
hp_df

[47]:

	id	label	lower respiratory tract	upper respiratory tract	trachea	bronchus	alveolus	lung	nose
0	CL:0002062	type I pneumocyte	1	0	0	0	1	1	0
1	CL:4033044	deuterosomal cell	0	0	0	0	0	0	0
2	CL:0000630	supporting cell	1	1	0	0	1	1	1
3	CL:0000064	ciliated cell	1	1	1	1	1	1	1
4	CL:0002145	ciliated columnar cell of tracheobronchial tree	1	0	0	1	0	1	0
5	CL:1000441	epithelial cell of viscerocranial mucosa	0	0	0	0	0	0	0
6	CL:0002638	bronchioalveolar stem cell	0	0	0	0	0	0	0
7	CL:0002139	endothelial cell of vascular tree	1	0	1	0	0	1	0
8	CL:0002184	basal proper cell of olfactory epithelium	0	0	0	0	0	0	0
9	CL:4028003	alveolar capillary type 2 endothelial cell	0	0	0	0	0	0	0
10	CL:0002329	basal epithelial cell of tracheobronchial tree	1	0	0	1	0	1	0
11	CL:0002167	olfactory epithelial cell	0	1	0	0	0	0	1
12	CL:0002076	endo-epithelial cell	1	0	0	1	0	1	0
13	CL:0002066	Feyrter cell	0	0	0	0	0	0	0
14	CL:0000076	squamous epithelial cell	1	0	1	0	1	1	1
15	CL:4030024	hillock cell	0	0	0	0	0	0	0
16	CL:0017000	pulmonary ionocyte	1	0	0	1	0	1	0
17	CL:4040003	fetal pre-type II pneumocyte	0	0	0	0	0	0	0
18	CL:0002171	globose cell of olfactory epithelium	0	0	0	0	0	0	0
19	CL:0000160	goblet cell	1	0	0	1	0	1	0
20	CL:4033023	airway submucosal gland collecting duct epithe...	0	0	0	0	0	0	0
21	CL:4033048	respiratory suprabasal cell	0	0	0	0	0	0	0
22	CL:0000066	epithelial cell	1	1	1	1	1	1	1
23	CL:0000165	neuroendocrine cell	1	0	0	1	0	1	0
24	CL:0000322	pneumocyte	1	0	0	0	1	1	0
25	CL:0002633	respiratory basal cell	0	0	0	0	0	0	0
26	CL:0002330	undifferentiated cell of bronchus epithelium	0	0	0	0	0	0	0
27	CL:0002209	intermediate epitheliocyte	1	0	0	1	0	1	0
28	CL:0002063	type II pneumocyte	1	0	0	0	1	1	0
29	CL:0000319	mucus secreting cell	1	0	1	1	0	1	0
30	CL:0002204	brush cell	1	0	0	1	0	1	0
31	CL:4033037	mucus secreting cell of tracheobronchial tree ...	0	0	0	0	0	0	0
32	CL:4028002	alveolar capillary type 1 endothelial cell	0	0	0	0	0	0	0
33	CL:0000158	club cell	1	0	0	1	0	1	0
34	CL:0000077	mesothelial cell	1	0	0	0	0	0	0
35	CL:0000185	myoepithelial cell	0	0	0	0	0	0	0

[48]:

hp_df.set_index('label', inplace=True)

plt.figure(figsize=(10, 8))
sns.heatmap(hp_df.iloc[:, 1:], cmap='viridis', cbar=True, linewidths=.5)
plt.xlabel('Anatomical Entities')
plt.ylabel('Cell Types')
plt.title('Heatmap of Dissected Anatomical Entities')
plt.show()

../../../_images/examples_Ontologies_CL_Lung-Epithelial-Cell-Types_39_0.png

[53]:

hp_only_df = hp_df.astype(bool) & ~xp_df.astype(bool)
hp_only_df

[53]:

	id	lower respiratory tract	upper respiratory tract	trachea	bronchus	alveolus	lung	nose
label
type I pneumocyte	False	False	False	False	False	False	False	False
deuterosomal cell	False	False	False	False	False	False	False	False
supporting cell	False	True	False	False	False	True	True	False
ciliated cell	False	False	True	True	False	True	True	True
ciliated columnar cell of tracheobronchial tree	False	False	False	False	False	False	True	False
epithelial cell of viscerocranial mucosa	False	False	False	False	False	False	False	False
bronchioalveolar stem cell	False	False	False	False	False	False	False	False
endothelial cell of vascular tree	False	False	False	True	False	False	False	False
basal proper cell of olfactory epithelium	False	False	False	False	False	False	False	False
alveolar capillary type 2 endothelial cell	False	False	False	False	False	False	False	False
basal epithelial cell of tracheobronchial tree	False	False	False	False	False	False	True	False
olfactory epithelial cell	False	False	False	False	False	False	False	False
endo-epithelial cell	False	False	False	False	False	False	False	False
Feyrter cell	False	False	False	False	False	False	False	False
squamous epithelial cell	False	False	False	True	False	False	False	True
hillock cell	False	False	False	False	False	False	False	False
pulmonary ionocyte	False	False	False	False	True	False	False	False
fetal pre-type II pneumocyte	False	False	False	False	False	False	False	False
globose cell of olfactory epithelium	False	False	False	False	False	False	False	False
goblet cell	False	False	False	False	False	False	False	False
airway submucosal gland collecting duct epithelial cell	False	False	False	False	False	False	False	False
respiratory suprabasal cell	False	False	False	False	False	False	False	False
epithelial cell	False	False	False	False	False	False	False	False
neuroendocrine cell	False	False	False	False	False	False	False	False
pneumocyte	False	False	False	False	False	False	False	False
respiratory basal cell	False	False	False	False	False	False	False	False
undifferentiated cell of bronchus epithelium	False	False	False	False	False	False	False	False
intermediate epitheliocyte	False	False	False	False	True	False	True	False
type II pneumocyte	False	False	False	False	False	False	False	False
mucus secreting cell	False	False	False	False	False	False	False	False
brush cell	False	False	False	False	False	False	False	False
mucus secreting cell of tracheobronchial tree submucosal gland	False	False	False	False	False	False	False	False
alveolar capillary type 1 endothelial cell	False	False	False	False	False	False	False	False
club cell	False	False	False	False	True	False	True	False
mesothelial cell	False	False	False	False	False	False	False	False
myoepithelial cell	False	False	False	False	False	False	False	False

[55]:

#hp_only_df.set_index('label', inplace=True)

plt.figure(figsize=(10, 8))
sns.heatmap(hp_only_df.iloc[:, 1:], cmap='viridis', cbar=True, linewidths=.5)
plt.xlabel('Anatomical Entities')
plt.ylabel('Cell Types')
plt.title('Heatmap of Dissected Anatomical Entities')
plt.show()

../../../_images/examples_Ontologies_CL_Lung-Epithelial-Cell-Types_41_0.png

[ ]: