UMAP Visualization

UMAP_viz.py

import umap
import zipfile
import urllib.request
import pandas as pd
import umap
urllib.request.urlretrieve("https://ftp.ncbi.nlm.nih.gov/geo/samples/GSM2230nnn/GSM2230760/suppl/GSM2230760_human4_umifm_counts.csv.gz", 'GSM2230760_human4_umifm_counts.csv.gz')

data = pd.read_csv('GSM2230760_human4_umifm_counts.csv.gz')
data.drop(columns='Unnamed: 0', inplace=True)
data.set_index("barcode", inplace=True)

reducer = umap.UMAP()
embedding = reducer.fit_transform(data.loc[:, data.columns != 'assigned_cluster'].values)

import matplotlib.pyplot as plt
import seaborn as sn

print(data.assigned_cluster.unique())

# Map the cell types to categorical variable (integer number) for color coding
clr = data.assigned_cluster.map({"ductal":0, "delta":1, "alpha":2,   "activated_stellate": 3,   "beta": 4, "macrophage": 5, "quiescent_stellate": 6, "acinar":7,  "gamma":8, "epsilon":9, "endothelial":10, "t_cell":11, "mast":12, "schwann":13})

plt.rcParams['figure.figsize'] = [26.0, 24.0]
plt.scatter(
    embedding[:, 0],
    embedding[:, 1], s =10, 
    c=clr, cmap="magma")