-
Notifications
You must be signed in to change notification settings - Fork 0
/
enrichmenTE_cluster.py
248 lines (206 loc) · 6.84 KB
/
enrichmenTE_cluster.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
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
#!/usr/bin/env python3
import os
import subprocess
import glob
import pandas as pd
def mkdir(dir):
if os.path.isdir(dir):
print("Directory %s exists" % dir)
return
try:
os.mkdir(dir)
except OSError:
print("Creation of the directory %s failed" % dir)
else:
print("Successfully created the directory %s " % dir)
def merge_bed(bed_list, bed_out):
with open(bed_out, "w") as output:
for bed in bed_list:
sample_name = os.path.basename(bed).replace(".nonref.bed", "")
with open(bed, "r") as input:
for line in input:
entry = line.replace("\n", "").split("\t")
chrom = entry[0]
start = entry[1]
end = entry[2]
family = entry[3]
score = entry[4]
strand = entry[5]
info = "|".join([family, sample_name])
out_line = "\t".join([chrom, start, end, info, score, strand])
output.write(out_line + "\n")
def filter_region_bed(bed_in, region_filter, bed_out):
with open(bed_out, "w") as output:
subprocess.call(
["bedtools", "intersect", "-a", bed_in, "-b", region_filter, "-u"],
stdout=output,
)
def filter_family_bed(bed_in, family_filter, bed_out, method):
families = set(family_filter.split(","))
with open(bed_out, "w") as output, open(bed_in, "r") as input:
for line in input:
entry = line.replace("\n", "").split("\t")
family = entry[3].split("|")[0]
if method == "include":
if family in families:
output.write(line)
else:
if family not in families:
output.write(line)
def sort_bed(bed_in, bed_out):
with open(bed_out, "w") as output:
subprocess.call(["bedtools", "sort", "-i", bed_in], stdout=output)
def cluster_bed(bed_in, bed_out):
window = 0
with open(bed_out, "w") as output:
subprocess.call(
["bedtools", "cluster", "-s", "-d", str(window), "-i", bed_in],
stdout=output,
)
def get_te_info(meta):
te_info_dict = dict()
with open(meta, "r") as input:
for line in input:
entry = line.replace("\n", "").split("\t")
te_info_dict[entry[0]] = entry[1]
return te_info_dict
def get_method_bed(
te_out_dirs,
filter_region,
outdir,
include_families,
exclude_families,
exclude_samples,
prefix,
):
output_prefix = prefix
pattern = "/**/detect/*nonref.bed"
bed_list = []
for te_out_dir in te_out_dirs:
bed_files = glob.glob(te_out_dir + pattern, recursive=True)
bed_list = bed_list + bed_files
# exclude samples
if exclude_samples is not None:
exclude_sample_list = exclude_samples.replace(" ", "").split(",")
new_bed_list = []
for bed in bed_list:
include = True
for exclude_sample in exclude_sample_list:
if exclude_sample in bed:
include = False
if include:
new_bed_list.append(bed)
bed_list = new_bed_list
# keep only nonref, merge per method
bed_merged = os.path.join(outdir, output_prefix + ".merge.bed")
merge_bed(bed_list=bed_list, bed_out=bed_merged)
# filter by region
bed_filtered = os.path.join(outdir, output_prefix + ".filter.bed")
filter_region_bed(
bed_in=bed_merged,
region_filter=filter_region,
bed_out=bed_filtered,
)
# filter by family
if include_families is not None:
bed_filtered_tmp = bed_filtered + ".tmp"
filter_family_bed(
bed_in=bed_filtered,
family_filter=include_families,
bed_out=bed_filtered_tmp,
method="include",
)
os.rename(bed_filtered_tmp, bed_filtered)
if exclude_families is not None:
bed_filtered_tmp = bed_filtered + ".tmp"
filter_family_bed(
bed_in=bed_filtered,
family_filter=exclude_families,
bed_out=bed_filtered_tmp,
method="exclude",
)
os.rename(bed_filtered_tmp, bed_filtered)
# sort and cluster, use method specific window
bed_sort = os.path.join(outdir, output_prefix + ".sort.bed")
sort_bed(bed_in=bed_filtered, bed_out=bed_sort)
bed_cluster = os.path.join(outdir, output_prefix + ".cluster.bed")
cluster_bed(bed_in=bed_sort, bed_out=bed_cluster)
# os.remove(bed_merged)
# os.remove(bed_filtered)
# os.remove(bed_sort)
return bed_cluster
def bed2matrix(bed, outgroup):
# from clustered bed to binary data matrix
header = [
"chr",
"start",
"end",
"info",
"score",
"strand",
"cluster",
]
df = pd.read_csv(bed, delimiter="\t", names=header)
info = df["info"].str.split("|", expand=True)
df["family"] = info[0]
df["sample"] = info[1]
df.drop(["info"], inplace=True, axis=1)
## filter out clusters with where a sample appears > 1 time
df.drop_duplicates(subset=["sample", "cluster"], keep=False, inplace=True)
## filter out clusters with more than two TE families
df = df.groupby("cluster").filter(lambda x: x["family"].nunique() == 1)
# convert to matrix
df["value"] = 1
matrix = df.pivot_table(
index="sample", columns="cluster", values="value", fill_value=0
)
if outgroup is not None and outgroup != "":
root_row = pd.Series(0, index=matrix.columns)
root_row.name = outgroup
matrix = matrix.append(root_row)
return matrix
def get_cluster(input, outdir, prefix):
tree_file = os.path.join(outdir, prefix + ".nwk")
pdf_file = os.path.join(outdir, prefix + ".pdf")
subprocess.call(
["Rscript", "--vanilla", "get_cluster.R", input, tree_file, pdf_file]
)
return tree_file, pdf_file
def cluster(
prefix,
enrichmente_out_dirs,
filter_region,
outgroup,
out,
include_families,
exclude_families,
exclude_samples,
):
# args = get_args()
abspath = os.path.abspath(__file__)
dname = os.path.dirname(abspath)
os.chdir(dname)
# get clustered bed from each matrix
bed_cluster = get_method_bed(
enrichmente_out_dirs,
filter_region,
out,
include_families,
exclude_families,
exclude_samples,
prefix,
)
# generate data matrix
matrix = bed2matrix(bed_cluster, outgroup=outgroup)
# write matrix to csv file
matrix_file = os.path.join(out, prefix + ".matrix.csv")
matrix.to_csv(
matrix_file,
sep=",",
index=True,
header=True,
)
# generate NJ tree
tree_file, pdf_file = get_cluster(matrix_file, out, prefix)
return tree_file, pdf_file
# main()