-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathCUTnRUN_PCA.py
150 lines (108 loc) · 5.06 KB
/
CUTnRUN_PCA.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
#!/usr/bin/env python
# coding: utf-8
# In[101]:
import sys
import numpy as np
import pandas as pd
import seaborn as sb
import matplotlib
import matplotlib.pyplot as plt
pd.options.mode.chained_assignment = None # stop annoying warning when clipping columns. default='warn'
from combat.pycombat import pycombat
# In[97]:
# prepare data
# the indexes correspond to the gene names
# the column names correspond to the sample names
# CUT n RUN
input_df = pd.read_csv("/Users/jra/Desktop/CTCF_tmp/data_filtered_sort_limma.txt",
sep='\t', index_col=0)
# HiChIP 1D
#input_df = pd.read_csv("/Users/jra/Dropbox/3dname/28_feb_2023/E14_D0-4_WT-TKO_CTCF_HiChIP_rep1-2_R1-2_trimV1_mm10_macs2_broad_peaks_DNAme_filtered_1CpG_data_named.txt",
# sep='\t', index_col=0)
for column_name in input_df.columns:
print(column_name)
print(input_df.shape)
#print(df_expression.columns)
#print(df_expression.head(2))
# In[98]:
# CUT n RUN
df = input_df[["E14_D0_WT_CTCF_CUTnRUN_AMS042021_rep1_crop36_trimV5_mm10.bam_RPKM",
"E14_D0_WT_CTCF_CUTnRUN_AMS042021_rep2_crop36_trimV5_mm10.bam_RPKM",
"E14_D0_TKO_CTCF_CUTnRUN_AMS042021_rep1_crop36_trimV5_mm10.bam_RPKM",
"E14_D0_TKO_CTCF_CUTnRUN_AMS042021_rep2_crop36_trimV5_mm10.bam_RPKM",
"E14_D4_WT_CTCF_CUTnRUN_AMS042021_rep1_crop36_trimV5_mm10.bam_RPKM",
"E14_D4_WT_CTCF_CUTnRUN_AMS042021_rep2_crop36_trimV5_mm10.bam_RPKM",
"E14_D4_TKO_CTCF_CUTnRUN_AMS042021_rep1_crop36_trimV5_mm10.bam_RPKM",
"E14_D4_TKO_CTCF_CUTnRUN_AMS042021_rep2_crop36_trimV5_mm10.bam_RPKM"]]
# HiChIP 1D
#df = input_df[["E14_D4_WT_CTCF_HiChIP_deep_shallow_AMS112022_rep1_R1-2_trimV1_mm10.bam_RPKM",
# "E14_D0_WT_CTCF_HiChIP_deep_shallow_AMS112022_rep1_R1-2_trimV1_mm10.bam_RPKM",
# "E14_D4_TKO_CTCF_HiChIP_deep_shallow_AMS112022_rep1_R1-2_trimV1_mm10.bam_RPKM",
# "E14_D0_TKO_CTCF_HiChIP_deep_shallow_AMS112022_rep1_R1-2_trimV1_mm10.bam_RPKM",
# "E14_D4_TKO_CTCF_HiChIP_AMS012023_rep2_R1-2_trimV1_mm10.bam_RPKM",
# "E14_D0_WT_CTCF_HiChIP_AMS012023_rep2_R1-2_trimV1_mm10.bam_RPKM",
# "E14_D4_WT_CTCF_HiChIP_AMS012023_rep2_R1-2_trimV1_mm10.bam_RPKM",
# "E14_D0_TKO_CTCF_HiChIP_AMS012023_rep2_R1-2_trimV1_mm10.bam_RPKM"]]
df
#df_expression
# SOMETHING IS WRONG WITH THE TABLE LETS FIND OUT WHAT
# df_expression.isnull().values.any()
# df_expression.isnull().sum()
# null_data = df_expression[df_expression.isnull().any(axis=1)]
# null_data
# should return 0 rows if the table is nice
df = df.loc[(df!=0).any(1)] # drop rows that have all 0 values
print("Dropping rows with only 0 values")
print(df.shape)
df
# In[99]:
matplotlib.rcParams['svg.fonttype'] = 'none' # saves fonts as a text object instead of a vector path
plt.rcParams['font.size'] = 8
plt.rcParams["figure.figsize"] = (20,4)
# plot the distribution of log2(RPKM+1) gene expression values (uncorrected for batch effects)
boxPlot = plt.boxplot(np.log2(df.clip(upper=None, lower=0)+1), \
labels=df.columns, \
patch_artist=True, notch=True, \
medianprops={'color':'Black'}, \
flierprops={'markersize':2, 'marker':'o', 'markerfacecolor':'white', 'markeredgecolor':'black', 'markeredgewidth':0.5})
plt.xticks(rotation=90)
plt.ylim(-2.1, None)
plt.show() # showing the figure "moves" the figure to Shell, leaving nothing behind. Show after saving file.
# In[100]:
from sklearn.decomposition import PCA
samples = df.transpose()
#samples = df
pca = PCA(2)
projected = pca.fit_transform(samples)
print(df.shape)
print(samples.shape)
print(projected.shape)
pca1variation = (pca.explained_variance_ratio_ * 100)[0].round(2)
pca2variation = (pca.explained_variance_ratio_ * 100)[1].round(2)
# use the explained variance to size the output graph
plt.rcParams["figure.figsize"] = ((pca1variation/4).astype(int), (pca2variation/4).astype(int))
plt.scatter(projected[:, 0], projected[:, 1],
edgecolor='none', alpha=1, s=200)
x_title = 'PC1: ' + pca1variation.astype(str) + '% of variation explained'
y_title = 'PC2: ' + pca2variation.astype(str) + '% of variation explained'
plt.xlabel(x_title)
plt.ylabel(y_title)
plt.title('made by Julien using python in a jupyter notebook')
for i, label in enumerate(df.columns):
plt.annotate(label, (projected[:,0][i], projected[:,1][i]))
outFigure = "/Users/jra/Desktop/PCA_plot_PCscaled.svg"
plt.savefig(outFigure)
plt.close()
plt.rcParams["figure.figsize"] = (4,4)
plt.scatter(projected[:, 0], projected[:, 1],
edgecolor='none', alpha=1, s=200)
x_title = 'PC1: ' + pca1variation.astype(str) + '% of variation explained'
y_title = 'PC2: ' + pca2variation.astype(str) + '% of variation explained'
plt.xlabel(x_title)
plt.ylabel(y_title)
plt.title('made by Julien using python in a jupyter notebook')
for i, label in enumerate(df.columns):
plt.annotate(label, (projected[:,0][i], projected[:,1][i]))
outFigure = "/Users/jra/Desktop/PCA_plot.svg"
plt.savefig(outFigure)
#plt.close()