This repository has been archived by the owner on Jul 1, 2024. It is now read-only.
-
Notifications
You must be signed in to change notification settings - Fork 1
/
unstructured.py
198 lines (170 loc) · 10.5 KB
/
unstructured.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
#!/usr/bin/env python
# coding: utf-8
from sarscov2_util import *
import os
import argparse
parser = argparse.ArgumentParser(description='Get conserved, unpaired stretches using Contrafold 2.0 base-pairing probabilities and RNAplfold. Output file results/unpaired_betacov_overlaps.csv has intervals that are unpaired at the specified thresholds and conserved in SARS-related sequences. Output file results/sarscov2_conserved_unstructured.csv has intervals that are unpaired and conserved in SARS-CoV-2 sequences. Output file results/unpaired_intervals.bed has all unpaired intervals at the specified threshold regardless of conservation. Output file results/rnaplfold_intervals.csv has unpaired intervals that overlap between Contrafold 2.0 and RNAplfold.')
parser.add_argument('--sarsr_aln_file', default='alignments/RR_nCov_alignment2_021120_muscle.fa', help="Alignment file for SARS-related sequences in fasta format")
parser.add_argument('--sarscov2_aln_file', default='alignments/gisaid_mafft_ncbi.fa', help="Alignment file for SARS-CoV-2 sequences in fasta format")
parser.add_argument('--min_len', type=int, default=15, help="Minimum length of reported conserved unpaired intervals")
parser.add_argument('--sarsr_conservation', type=float, default=0.9, help="Required conservation level for intervals in SARS-related sequences (float from 0 to 1)")
parser.add_argument('--sarscov2_conservation', type=float, default=0.97, help="Required conservation level for intervals in SARS-CoV-2 sequences (float from 0 to 1)")
parser.add_argument('--min_unpaired_prob', type=float, default=0.6, help="Minimum probability unpaired for the position to be classified as unstructured; all positions in unpaired intervals must have unpaired probabilities above this threshold.")
parser.add_argument('--do_significance_tests', dest='do_significance_tests', action='store_true', help="If specified, will do significance tests for the overlap of SARS-related conserved intervals and SARS-CoV-2 conserved intervals")
args = parser.parse_args()
aln_file_1 = args.sarsr_aln_file
sarscov2_aln_file = args.sarscov2_aln_file
MIN_SIZE = args.min_len
SARSR_CONSERVATION = args.sarsr_conservation
SARSCOV2_CONSERVATION = args.sarscov2_conservation
MIN_UNPAIRED = args.min_unpaired_prob
do_significance_tests = args.do_significance_tests
def get_unpaired_probs(int_start, int_end, ref_seq):
sequence = ref_seq[int_start:int_end]
bp_matrix = bpps(sequence, package='contrafold_2')
paired_probs = np.sum(bp_matrix, axis=0)
return 1 - paired_probs
# Just as in RNAz, use 120 bp windows and slide 40 bp
def get_per_position_unpaired(ref_seq, window_size=120, slide=40, print_freq=1000):
total_unpaired_probs = np.zeros(len(ref_seq))
num_unpaired_probs = np.zeros(len(ref_seq))
ii = 0
while ii < len(ref_seq):
window_len = min(window_size, len(ref_seq) - ii)
if ii % print_freq == 0:
print("Window: %d to %d" % (ii, ii + window_len))
unpaired_probs = get_unpaired_probs(ii, ii + window_len, ref_seq)
total_unpaired_probs[ii:(ii + window_len)] += unpaired_probs
num_unpaired_probs[ii:(ii + window_len)] += np.ones(window_len)
ii += slide
return total_unpaired_probs/num_unpaired_probs
def get_unpaired_intervals(probs_unpaired, unpaired_cutoff=0.9, min_size=15):
intervals = []
start_int = -1
end_int = -1
in_int = False
int_size = 0
for ii in range(len(probs_unpaired)):
if (probs_unpaired[ii] < unpaired_cutoff):
if in_int:
if int_size > min_size:
avg_unpaired = np.mean(np.array(probs_unpaired[start_int:end_int]))
min_unpaired = min(probs_unpaired[start_int:end_int])
intervals += [(start_int + 1, end_int, avg_unpaired, min_unpaired)]
int_size = 0
in_int = False
else:
if not in_int:
in_int = True
start_int = ii
int_size += 1
end_int = ii
if in_int:
if int_size > min_size:
avg_unpaired = np.mean(np.array(probs_unpaired[start_int:end_int]))
min_unpaired = min(probs_unpaired[start_int:end_int])
intervals += [(start_int + 1, end_int, avg_unpaired, min_unpaired)]
return intervals
def write_unpaired_betacov(per_position_unpaired, ref_seq):
betacov_intervals = get_ref_intervals_from_file(aln_file_1, cutoff=SARSR_CONSERVATION, MIN_SIZE=MIN_SIZE)
unpaired_intervals = get_unpaired_intervals(per_position_unpaired, unpaired_cutoff=MIN_UNPAIRED, min_size=MIN_SIZE)
overlap_intervals = get_interval_overlap_size(unpaired_intervals, betacov_intervals, min_size=MIN_SIZE)
overlap_intervals = np.array(overlap_intervals)
p_vals = np.array([x[3] for x in overlap_intervals])
top_overlaps = overlap_intervals[np.argsort(p_vals)[::-1]]
csv_file = 'results/unpaired_betacov_overlaps.csv'
f = open(csv_file, 'w')
f.write('%s,%s,%s,%s,%s\n' % ("Start index", "end index", "average unpaired probability in interval", "minimum unpaired probability in interval", "interval sequence"))
for x in top_overlaps:
f.write('%d,%d,%f,%f,%s\n' % (int(x[0]), int(x[1]-1), float(x[2]), float(x[3]), ref_seq[(int(x[0])-1):int(x[1]-1)]))
f.close()
def write_unpaired_sarscov2(per_position_unpaired, ref_seq):
sarscov2_intervals = get_ref_intervals_from_file(sarscov2_aln_file, cutoff=SARSCOV2_CONSERVATION, MIN_SIZE=MIN_SIZE)
unpaired_intervals = get_unpaired_intervals(per_position_unpaired, unpaired_cutoff=MIN_UNPAIRED, min_size=MIN_SIZE)
overlap_intervals = get_interval_overlap_size(unpaired_intervals, sarscov2_intervals, min_size=MIN_SIZE)
overlap_intervals = np.array(overlap_intervals)
p_vals = np.array([x[3] for x in overlap_intervals])
top_overlaps = overlap_intervals[np.argsort(p_vals)[::-1]]
csv_file = 'results/sarscov2_conserved_unstructured.csv'
f = open(csv_file, 'w')
f.write('%s,%s,%s,%s,%s\n' % ("Start index", "end index", "average unpaired probability in interval", "minimum unpaired probability in interval", "interval sequence"))
for ii, x in enumerate(top_overlaps):
f.write('SARS-CoV-2-conserved-unstructured-%d,%d-%d,%f,%f,%s\n' % (ii+1,int(x[0]), int(x[1]-1), float(x[2]), float(x[3]), ref_seq[(int(x[0])-1):int(x[1]-1)]))
f.close()
def write_unpaired_intervals(unpaired_intervals):
# Print intervals in rank order of MCC secondary structure predictions
f = open('results/unpaired_intervals.bed', 'w')
for interval in unpaired_intervals:
f.write('NC_045512.2\t%d\t%d\tint\t0\t.\n' % (interval[0], interval[1]))
f.close()
def write_rnaplfold_intervals(intervals):
f = open('results/rnaplfold_intervals.csv', 'w')
f.write('%s,%s\n' % ("Start index", "end index"))
for interval in intervals:
f.write("%d, %d\n" % (interval[0], interval[1]))
f.close()
def get_rnaplfold_intervals(RNAplfold_filename, int_size=15, cutoff=0.25):
f = open(RNAplfold_filename)
rnaplfold_lines = f.readlines()
f.close()
rnaplfold_intervals = []
for rnaplfold_line in rnaplfold_lines[(int_size+2):]:
rnaplfold_items = rnaplfold_line.split('\t')
int_start = int(rnaplfold_items[0])
if float(rnaplfold_items[-1]) > cutoff:
last_int = (-1, -1)
if len(rnaplfold_intervals) > 0:
last_int = rnaplfold_intervals[-1]
# Continuation of last interval or not?
if int_start < last_int[1]:
rnaplfold_intervals = rnaplfold_intervals[:-1]
rnaplfold_intervals += [(last_int[0], int_start + int_size)]
else:
rnaplfold_intervals += [(int_start, int_start+int_size)]
print([x[0] for x in rnaplfold_intervals])
return rnaplfold_intervals
def get_rnaplfold_per_position_unpaired(RNAplfold_filename):
f = open(RNAplfold_filename)
rnaplfold_lines = f.readlines()
f.close()
rnaplfold_unpaired_probabilities = []
for rnaplfold_line in rnaplfold_lines[2:]:
rnaplfold_unpaired_probabilities += [float(rnaplfold_line.split('\t')[1])]
return rnaplfold_unpaired_probabilities
if __name__ == "__main__":
print("Getting and recording per-position unpaired")
sequences = get_sequences(aln_file_1)
(full_ref_seq, ref_seq) = get_ref_seq(sequences)
probs_unpaired_file = "results/probs_unpaired.txt"
per_position_unpaired = []
if os.path.isfile(probs_unpaired_file):
f = open(probs_unpaired_file)
per_position_unpaired = [float(x) for x in f.readlines()]
f.close()
else:
per_position_unpaired = get_per_position_unpaired(ref_seq)
f = open(probs_unpaired_file, 'w')
for cur_val in per_position_unpaired:
f.write("%f\n" % cur_val)
f.close()
unpaired_intervals = get_unpaired_intervals(per_position_unpaired, unpaired_cutoff=MIN_UNPAIRED, min_size=MIN_SIZE)
rnaplfold_unpaired_probabilities = get_rnaplfold_per_position_unpaired("RNAplfold/SARSCOV2_lunp")
rnaplfold_intervals = get_unpaired_intervals(rnaplfold_unpaired_probabilities, unpaired_cutoff=MIN_UNPAIRED, min_size=MIN_SIZE)
overlap_intervals = get_interval_overlap_size(unpaired_intervals, rnaplfold_intervals, min_size=5)
print("%d/%d Contrafold 2.0 intervals overlap with RNAplfold" % (len(overlap_intervals), len(unpaired_intervals)))
if do_significance_tests:
num_overlaps = get_num_overlaps_rnd_trials_size(unpaired_intervals, rnaplfold_intervals, len(ref_seq), min_size=5)
print("P-value for overlap: %.2E\n" % (np.sum(np.array(num_overlaps) >= len(overlap_intervals))/len(num_overlaps)))
overlap_intervals = get_interval_overlap_size(rnaplfold_intervals, unpaired_intervals, min_size=5)
print("%d/%d RNAplfold intervals overlap with Contrafold 2.0" % (len(overlap_intervals), len(rnaplfold_intervals)))
if do_significance_tests:
num_overlaps = get_num_overlaps_rnd_trials_size(rnaplfold_intervals, unpaired_intervals, len(ref_seq), min_size=5)
print("P-value for overlap: %.2E\n" % (np.sum(np.array(num_overlaps) >= len(overlap_intervals))/len(num_overlaps)))
print("Recording RNAplfold intervals that overlap with Contrafold 2.0 intervals")
write_rnaplfold_intervals(overlap_intervals)
print("Recording unpaired intervals conserved in SARS-related sequences")
write_unpaired_betacov(per_position_unpaired, ref_seq)
print("Recording unpaired intervals conserved in SARS-CoV-2 sequences")
write_unpaired_sarscov2(per_position_unpaired, ref_seq)
print("Recording unpaired intervals")
write_unpaired_intervals(unpaired_intervals)