-
Notifications
You must be signed in to change notification settings - Fork 3
/
pruner.py
229 lines (180 loc) · 8.07 KB
/
pruner.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
import Bio.Phylo as bp
import sys
import rdflib
class Prunable:
def get_subtree(self, contains=[], contains_ids=[], tree_uri=None,
format='newick', prune=True, filter=None, taxonomy=None,
handle=None):
# TODO: filter is not being used. Use cql.py to parse the query, then convert the
# requirements into SPARQL.
filter = self.validate_filter(filter)
if not contains or contains_ids: raise Exception('A list of taxa or ids is required.')
if tree_uri:
tree_uri = self.uri_from_id(tree_uri)
else:
trees = self.list_trees_containing_taxa(contains=contains,
show_counts=False,
filter=filter)
try:
tree_uri = trees.next()
except StopIteration:
raise Exception("An appropriate tree for this query couldn't be found.")
tree = self.subtree(list(contains), tree_uri,
taxonomy=taxonomy, prune=prune)
return self.serialize_trees(trees=[tree], format=format, handle=handle)
def find_mrca(self, taxa, graph, taxonomy=None):
assert len(taxa) > 0
cursor = self.get_cursor()
mrca = None
for n, taxon in enumerate(taxa[:]):
try:
result = self.find_name(graph, taxon, taxonomy)
if len(result) == 2: node_id, taxon, synonym = result + (None,)
else: node_id, taxon, synonym = result
ancestors = self.get_ancestors(graph, node_id)
if synonym: taxa[n] = synonym
except:
taxa[n] = None
continue
if not mrca:
mrca_ancestors = []
for (ancestor,) in ancestors:
mrca_ancestors.append(ancestor)
if not mrca_ancestors:
taxa[n] = None
continue
mrca = mrca_ancestors[0]
continue
for (ancestor,) in ancestors:
if ancestor in mrca_ancestors:
mrca = ancestor
mrca_ancestors = mrca_ancestors[mrca_ancestors.index(ancestor):]
break
if not mrca: raise Exception('None of these taxa are members of this tree.')
return mrca
def subtree(self, taxa, graph, taxonomy=None, prune=False):
'''Get a subtree containing a given set of taxa.'''
if taxa:
old_taxa = taxa[:]
mrca = self.find_mrca(taxa, graph, taxonomy)
# these taxa were changed by the MRCA query; they're either None (couldn't
# be found) or the name of a synonym
replace = {new:old for (new, old) in zip(taxa, old_taxa)
if new and old != new}
else:
mrca, replace = None, None
cursor = self.get_cursor()
query = '''sparql
PREFIX obo: <http://purl.obolibrary.org/obo/>
PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
SELECT DISTINCT ?n ?length ?parent ?label
WHERE {
GRAPH <''' + graph + '''> {
?n obo:CDAO_0000200 ?tree .
?n a ?type .
''' + ((
"?n obo:CDAO_0000179 %s option(transitive, t_min(0), t_step('step_no') as ?steps) ."
% rdflib.URIRef(mrca).n3()
) if mrca else '') + '''
OPTIONAL { ?n obo:CDAO_0000187 [ rdfs:label ?label ] . }
OPTIONAL { ?n obo:CDAO_0000143 [ obo:CDAO_0000193 [ obo:CDAO_0000215 ?length ] ] . }
OPTIONAL { ?n obo:CDAO_0000179 ?parent . }
FILTER (?type = obo:CDAO_0000108 || ?type = obo:CDAO_0000026)
}
}''' + ('ORDER BY ?steps ?n' if mrca else 'ORDER BY ?n')
if self.verbose: print query
cursor.execute(query)
root = None
nodes = {}
stmts = cursor
for _ in range(2):
redo = []
for stmt in stmts:
node_id, edge_length, parent, label = stmt
# replace synonymous names from the phylogeny with names from the query
if replace and label in replace: label = replace[label]
# create a clade for each node, and store in a dictionary by URI
if not node_id in nodes:
clade = bp.CDAO.Clade(name=label, branch_length=float(edge_length) if edge_length else None)
nodes[node_id] = clade
# this is the root node if it has no parent or if it's the MRCA
if root is None and ((node_id == mrca) if mrca else (parent is None)):
root = nodes[node_id]
elif parent and (parent in nodes):
nodes[parent].clades.append(clade)
else:
redo.append(stmt)
stmts = redo
tree = bp.CDAO.Tree(root=root, rooted=True)
if prune: result = pruned_tree(tree, old_taxa)
else: result = tree
return result
def get_ancestors(self, graph, node_id):
'''Query to get all ancestors of a node, starting with the most recent.'''
cursor = self.get_cursor()
query = '''sparql
PREFIX obo: <http://purl.obolibrary.org/obo/>
PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
SELECT DISTINCT ?ancestor
WHERE {
GRAPH %s {
%s obo:CDAO_0000179 ?ancestor
option(transitive, t_direction 1, t_step('step_no') as ?steps,
t_min 0, t_max 10000)
}
}
ORDER BY ?steps
''' % (rdflib.URIRef(graph).n3(), rdflib.URIRef(node_id).n3())
if self.verbose: print query
cursor.execute(query)
results = cursor
return results
def find_name(self, graph, taxon, taxonomy=None):
'''If taxon is the name of a node in this graph, return it; otherwise,
return a synonym from `taxonomy` that matches a name in this graph.'''
cursor = self.get_cursor()
if taxon.split()[-1] == 'sp.':
# when species is unidentified, fall back to searching for the genus
taxon = ' '.join(taxon.split()[:-1])
query = '''sparql
PREFIX obo: <http://purl.obolibrary.org/obo/>
PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
PREFIX skos: <http://www.w3.org/2004/02/skos/core#>
SELECT ?t ?label ''' + ('?synonym' if taxonomy else '') + '''
WHERE {
{
GRAPH %s {
?t obo:CDAO_0000187 [ rdfs:label ?label ]
FILTER (?label = %s)
}
}''' % (rdflib.URIRef(graph).n3(), rdflib.Literal(taxon).n3())
if taxonomy: query += ''' UNION {
GRAPH %s { ?t obo:CDAO_0000187 [ rdfs:label ?synonym ] }
GRAPH %s {
?x obo:CDAO_0000187 [ ?l1 ?synonym ; ?l2 ?label ]
FILTER (?label = %s &&
?l1 in (rdfs:label, skos:altLabel) &&
?l2 in (rdfs:label, skos:altLabel))
}
}''' % (rdflib.URIRef(graph).n3(), rdflib.URIRef(taxonomy).n3(), rdflib.Literal(taxon).n3())
query += '\n}'
if self.verbose: print query
cursor.execute(query)
results = cursor
return results.next()
def pruned_tree(tree, contains):
def prune_clade(tree, clade, root=False):
keep_pruning = True
while keep_pruning:
keep_pruning = False
for child in clade.clades:
if prune_clade(tree, child):
keep_pruning = True
break
if not root and not (clade.name in contains) and len(clade.clades) <= 1:
tree.collapse(clade)
return True
else:
return False
prune_clade(tree, tree.root, True)
return tree