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
249
250
251
252
253
254
255
|
#!/usr/bin/python3
# -*- coding: utf-8 -*-
#
'''
Compaction of facts. Non ambiguous overlapping paths are compacted
'''
__author__ = "Pierre Peterlongo"
__email__ = "pierre.peterlongo@inria.fr"
import sys
# import getopt
import K3000_common as kc
# import sorted_list
import argparse
def is_subsequence(x,y,position_suffix):
pos_on_x = position_suffix
pos_on_y = 0
len_x = len(x)
len_y = len(y)
while True:
if pos_on_y == len_y:
return True # All y was read, it is included in x
if pos_on_x == len_x:
return False # All x was read, thus y is not included in x
if kc.allele_value(x[pos_on_x]) == kc.allele_value(y[pos_on_y]):
pos_on_x += 1
pos_on_y += 1
else:
pos_on_x += 1
def remove_y_subsequence_of_x(x_ref,facts):
''' remove all y that are subsequence of x
Do not care about distances.
Exemple 3_0,4_1,5_10 is a subsequence of 0_0,1_12,2_13,3_13,4_12,5_123,6_1
'''
if len(x_ref) == 1: return # as we removed strict equalities, no read can be included in a read of size one.
n = len(x_ref)
# print ("x",x)
for x in [x_ref, kc.get_reverse_fact(x_ref)]:
for position_suffix in range(0,n):
u = x[position_suffix]
Y = facts.get_lists_starting_with_given_prefix([u])
if x in Y: Y.remove(x)
for y in Y:
if len(y)+position_suffix <= n and is_subsequence(x,y,position_suffix):
facts.remove(y)
if not kc.is_palindromic(y):
facts.remove(kc.get_reverse_fact(y))
def remove_strict_inclusions(facts):
''' remove all facts strictly included in any other '''
n = len(facts)
# to_remove=[False for i in range(n)]
checked = 0
for fact in facts.traverse():
if checked%100 == 0: sys.stderr.write(" Removing inclusions, "+str(checked)+" checked. Size facts "+str(len(facts))+" %.2f"%(100*checked/n)+"%\r")
checked += 1
remove_y_subsequence_of_x(fact,facts)
sys.stderr.write(" Removing inclusions, "+str(checked)+" checked. Size facts "+str(len(facts))+" %.2f"%(100*checked/n)+"%\n")
return facts
def right_unique_extention(facts,fact):#, unitig_lengths,k,min_conflict_overlap):
''' return the unique possible right fact extension with the largest overlap
return None if no right extensions or if more than one possible non colinear extensions
The returned extension can be fact itself
'''
n = len(fact)
# **** Get the largest right overlap with fact ****
for len_u in range(n-1,0,-1):
u = fact[-len_u:]
Y = facts.get_lists_starting_with_given_prefix(u)
if len(Y) == 0: continue # No y starting with u
if len(Y) > 1: return None,len_u # More than one unique y starting with u, for instance y and y'. Knowing that y is not included in y' it means necessary that y and y' are not colinear and that x is thus not right extensible.
y = Y[0] # We found the largest y right overlapping fact.
# **** check that all other right extensions are collinear with y.
# get all y' such that LCSP(fact,y') in [1,len(u)-1]
# get all y' starting with a suffix of u
Y = []
starting_positions = []
for starting_suffix_position in range(1,len_u):
suffix_u = u[starting_suffix_position:]
others = facts.get_lists_starting_with_given_prefix(suffix_u)
if len(others) >0:
Y += others
starting_positions += [starting_suffix_position for zz in range(len(others))]
if len(starting_positions)>0 and not kc.colinear(y,Y,starting_positions): return None,len_u # more than a unique right extention for x.
return y,len_u # A unique maximal right extention for x (unless gready: a unique largest extention, smaller possible extentions under the gready threahold)
# not any right extention found.
return None,None
def fusion (facts,x):
'''Main function. For a given fact x, we find y that overlap x with the highest overlap, such that :
1/ there exists no other y' right overlapping x that is not collinear with y
2/ there exists no other x' left overlapping y that is not collinear with x
Once done, we compact x and y, and this is finished for x.
'''
y,len_u = right_unique_extention(facts,x) # Define, if exists, the unique y != x having the largest right overlap with x.
if y == None: return 0 # if no unique right extension, finished, x is not right extensible.
if y == x: return 0 # Do not compact x with itself, else, enter an infinite loop
y_ = kc.get_reverse_fact(y) # what are left extentions of y, we right extend its reverse complement.
if y_ == x: return 0 # Do not compact x with its own reverse complement.
xprime_, _ = right_unique_extention(facts,y_) # Define, if exists, the unique xprime_ (!= y_) having the largest right overlap with y_.
if xprime_ == None: return 0 # if no unique left extension of the unique right extention of x, finished, x is not right extensible.
# assert xprime_ == kc.get_reverse_fact(x), "X "+str(x)+" xprime_ "+str(xprime_)+" Y "+str(y)+" Y_"+str(y_)+"\n"
# ***** FUSION *****
# 1/ remove x and its reverse complement if not palindromic
# 2/ if y is not x (x==y is possible if x is repeated 2,2,2,2 for instance or if prefix = suffix (1,2,1 for instance)), remove y and its reverse complement if not palindromic
# 3/ create the new xy facts and add it (sorted fashion)
# isthere = facts.contains(debug_id_node) #DEBUG
# 1
facts.remove(x)
# ### DEBUG
# if isthere and not facts.contains(debug_id_node):
# sys.stderr.write("\n\n\nSUPPRESS X \n\n\n"+str(x)+" "+str(y)+"\n\n\n")
# sys.exit(0)
# ### END DEBUG
if not kc.is_palindromic(x):
facts.remove(kc.get_reverse_fact(x))
# ### DEBUG
# if isthere and not facts.contains(debug_id_node):
# sys.stderr.write("\n\n\nSUPPRESS X_ \n\n\n"+str(kc.get_reverse_fact(x))+"\n\n\n")
# sys.exit(0)
# ### END DEBUG
# 2
if x != y:
facts.remove(y)
# ### DEBUG
# if isthere and not facts.contains(debug_id_node):
# sys.stderr.write("\n\n\nSUPPRESS Y \n\n\n"+str(x)+" "+str(y)+"\n\n\n")
# sys.exit(0)
# ### END DEBUG
if not kc.is_palindromic(y):
facts.remove(kc.get_reverse_fact(y))
# ### DEBUG
# if isthere and not facts.contains(debug_id_node):
# sys.stderr.write("\n\n\nSUPPRESS Y_ \n\n\n"+str(x)+" "+str(y)+"\n\n\n")
# sys.exit(0)
# ### END DEBUG
# 3
new = x+y[len_u:]
facts.sorted_add(new)
if not kc.is_palindromic(new): facts.sorted_add(kc.get_reverse_fact(new))
# if isthere and not facts.contains(debug_id_node):
# sys.stderr.write("\n\n\n"+str(x)+" "+str(y)+" "+str(new)+"\n\n\n")
# sys.exit(0)
# we made a compaction, return 1.
return 1
def compaction(facts):
''' Compaction of all fact in facts
If a fact was not compacted, it will never be compacted.
If it was compacted, maybe it will be re-compacted later on. However, no need to re-run the fusion on this read as
- either I could have been compacted on the left and this was done before or this will be done latter or
- it will be right extended later: the 'new' (the compacted fact) fact is higher in the lexicographic order than the original fact (as it is longer), thus its position is higher in the facts data structure, thus it will be seen again later.
Note that this may be not true in parallel computations.
'''
checked = 0
compacted = 0
n = len(facts)
for fact in facts.traverse():
if checked%100 == 0:
sys.stderr.write(" Compacting, "+str(checked)+" checked. Size facts "+str(len(facts))+" %.2f"%(100*checked/n)+"%, "+str(compacted)+" couple of nodes compacted\r")
checked += 1
witness = fusion(facts,fact)
if witness == 1: # a fusion was done
compacted += 1
sys.stderr.write(" Compacting, "+str(checked)+" checked. Size facts "+str(len(facts))+" %.2f"%(100*checked/n)+"%, "+str(compacted)+" couple of nodes compacted\n")
return facts
def main():
'''
Compaction of set of facts coded as set of ids of unitigs
'''
parser = argparse.ArgumentParser(description='Compaction of set of facts coded as set of ids of unitigs.')
parser.add_argument("input_file", type=str,
help="input file containing dbg paths as a list of unitig ids, eg. on line looks like \"-1;24;198;\"" )
args = parser.parse_args()
input_file = str(args.input_file)
sys.stderr.write(" Load phased alleles \r")
facts, nb_nonvalid = kc.generate_facts_from_disco_pashing(input_file)
sys.stderr.write(" Load phased alleles. Done - nb facts="+ str(len(facts))+". "+str(nb_nonvalid)+" were removed because they are non valid\n")
sys.stderr.write(" Add reverse complements \r")
kc.add_reverse_facts(facts)
sys.stderr.write(" Add reverse complements. Done - nb facts="+ str(len(facts))+"\n")
sys.stderr.write(" Remove strict inclusions\r")
facts = remove_strict_inclusions(facts)
sys.stderr.write(" Remove strict inclusions. Done - nb facts="+ str(len(facts))+"\n")
sys.stderr.write(" Compaction of simple paths \r")
facts = compaction(facts)
sys.stderr.write(" Compaction of simple paths. Done - nb facts="+ str(len(facts))+"\n")
sys.stderr.write(" Remove2 strict inclusions\r")
facts = remove_strict_inclusions(facts)
sys.stderr.write(" Remove2 strict inclusions. Done - nb facts="+ str(len(facts))+"\n")
sys.stderr.write(" Compaction2 of simple paths \r")
facts = compaction(facts)
sys.stderr.write(" Compaction2 of simple paths. Done - nb facts="+ str(len(facts))+"\n")
sys.stderr.write(" Remove3 strict inclusions\r")
facts = remove_strict_inclusions(facts)
sys.stderr.write(" Remove3 strict inclusions. Done - nb facts="+ str(len(facts))+"\n")
sys.stderr.write(" Compaction3 of simple paths \r")
facts = compaction(facts)
sys.stderr.write(" Compaction3 of simple paths. Done - nb facts="+ str(len(facts))+"\n")
sys.stderr.write(" Print canonical compacted phased alleles\n")
kc.print_maximal_facts(facts)
if __name__ == "__main__":
main()
|
