scEnrichment¶

In this tutorial, we will provide a quick guideline for applying the scEnrichment feature of discotoolkit on DEGs (Differentially Expressed Genes). Follow these steps:

1. First, we load the DEGs from Example 1, available on the DISCO website.
2. Finally, we convert the retrieved data into a Pandas DataFrame, which will serve as the input for the dt.CELLiD_enrichment function. By running this function, we can obtain the desired enrichment analysis.

In [1]:

# for google colab
# pip install discotoolkit
# import package

import discotoolkit as dt
import scanpy as sc
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns

# adding ignore warning to clean the code
import warnings

# Ignore all warnings
warnings.filterwarnings('ignore')

%load_ext autoreload
%autoreload 2

sns.set_theme(rc={'figure.dpi': 300})

# for google colab # pip install discotoolkit # import package import discotoolkit as dt import scanpy as sc import pandas as pd import matplotlib.pyplot as plt import seaborn as sns # adding ignore warning to clean the code import warnings # Ignore all warnings warnings.filterwarnings('ignore') %load_ext autoreload %autoreload 2 sns.set_theme(rc={'figure.dpi': 300})

The user can input either a gene list or a gene list with log fold change as the input to the dt.CELLiD_enrichment function.

Example of the DEGs:

In [2]:

# testing DEG from DEGs of acinar cell(Peng, Junya et al.)
# Reference is in DISCO website

test_genes = {"gene":["PRSS1", "CTRB2", "CELA3A", "CTRB1", "REG1B", "CLPS", "CPB1", "CPA1", "PLA2G1B", "REG3A", "CTRC"],
            #   "fc":[5.236605052, 5.179753462, 4.724315075, 4.702706704, 4.65145949, 4.513887613, 4.351968886, 4.311988687, 4.272185339, 4.253882194, 4.208933992]
              }

deg_df = pd.DataFrame(test_genes)

deg_df.head()

# testing DEG from DEGs of acinar cell(Peng, Junya et al.) # Reference is in DISCO website test_genes = {"gene":["PRSS1", "CTRB2", "CELA3A", "CTRB1", "REG1B", "CLPS", "CPB1", "CPA1", "PLA2G1B", "REG3A", "CTRC"], # "fc":[5.236605052, 5.179753462, 4.724315075, 4.702706704, 4.65145949, 4.513887613, 4.351968886, 4.311988687, 4.272185339, 4.253882194, 4.208933992] } deg_df = pd.DataFrame(test_genes) deg_df.head()

Out[2]:

	gene
0	PRSS1
1	CTRB2
2	CELA3A
3	CTRB1
4	REG1B

Now that we have our data ready, we can proceed to run the function.

In [3]:

# get enrichment analysis result using dt.CELLiD_enrichment
results = dt.CELLiD_enrichment(deg_df)

# get enrichment analysis result using dt.CELLiD_enrichment results = dt.CELLiD_enrichment(deg_df)

INFO:root:Downloading gene set reference...

INFO:root:Comparing the ranked gene list to reference gene sets...

[Parallel(n_jobs=1)]: Done  40 tasks      | elapsed:    0.3s

[Parallel(n_jobs=1)]: Done 161 tasks      | elapsed:    0.9s

[Parallel(n_jobs=1)]: Done 364 tasks      | elapsed:    1.6s

[Parallel(n_jobs=1)]: Done 647 tasks      | elapsed:    2.9s

[Parallel(n_jobs=1)]: Done 1012 tasks      | elapsed:    4.3s

[Parallel(n_jobs=1)]: Done 1253 out of 1253 | elapsed:    5.5s finished

In [4]:

# get the top 10 results
results.head(10)

# get the top 10 results results.head(10)

Out[4]:

	pval	or	name	gene	background	overlap	geneset
9	0.0	700.394	Marker of Acinar cell in pancreas_cell	CTRC,CELA3A,REG1B,REG3A,CPB1,CLPS,CPA1,PRSS1,P...	4228	11	81
3	0.0	609.570	Marker of Acinar cell in PDAC_pancreas	CTRC,CELA3A,CPB1,CLPS,CPA1,PRSS1,PLA2G1B,CTRB2...	4105	11	89
22	0.0	317.643	Marker of Paneth cell in intestine	REG1B,REG3A	4493	2	43
13	0.0	278.933	Marker of Acinar cell in type_1_diabetes_pancreas	CTRC,CELA3A,REG1B,CPB1,CLPS,CPA1,PRSS1,PLA2G1B...	1104	11	55
0	0.0	159.589	Marker of Acinar cell in type_2_diabetes_pancreas	CTRC,CELA3A,REG1B,REG3A,CPB1,CLPS,CPA1,PRSS1,P...	1543	11	117
21	0.0	154.783	type 1 diabetes vs control for T cell in type_...	REG1B,PRSS1,REG3A,CTRB2,CLPS,CTRB1,CELA3A,PLA2...	1104	9	31
20	0.0	117.889	type 1 diabetes vs control for Perivascular ce...	REG3A,REG1B,PRSS1,CELA3A,CTRB2,CLPS,PLA2G1B,CP...	1104	9	38
16	0.0	109.330	type 1 diabetes vs control for Delta cell in t...	REG3A,REG1B,PRSS1,CLPS,CELA3A,PLA2G1B,CTRB1,CTRB2	1104	8	29
14	0.0	89.915	type 1 diabetes vs control for Alpha cell in t...	REG3A,REG1B,PRSS1,CTRB2,CELA3A,CTRB1,CLPS,PLA2...	1104	9	47
18	0.0	39.280	type 1 diabetes vs control for Fibroblast in t...	REG1B,REG3A,CTRB2,PRSS1,CTRB1,PLA2G1B,CPB1,CEL...	1104	10	143

We can visualize the results using a horizontal barplot with the help of the seaborn library.

In [5]:

# setting the theme to whitegrid
sns.set_theme(style="whitegrid")

# Initialize the matplotlib figure
f, ax = plt.subplots(figsize=(6, 15))

# plot only the top 20 results
sns.barplot(x="or", y="name", data=results.head(20), color="b")

# adding axis title
ax.set(ylabel="Gene sets",
       xlabel="Odds Ratio")

# make the plot look cleaner and nicer
sns.despine(left=True, bottom=True)

# setting the theme to whitegrid sns.set_theme(style="whitegrid") # Initialize the matplotlib figure f, ax = plt.subplots(figsize=(6, 15)) # plot only the top 20 results sns.barplot(x="or", y="name", data=results.head(20), color="b") # adding axis title ax.set(ylabel="Gene sets", xlabel="Odds Ratio") # make the plot look cleaner and nicer sns.despine(left=True, bottom=True)