|
0
|
1 import os.path
|
|
|
2 import common
|
|
|
3
|
|
|
4 class ReferenceBins:
|
|
|
5
|
|
|
6 def __init__(self, db_spec_path = None):
|
|
|
7 """
|
|
|
8 @param db_spec_path string path to fasta databases specification file. This file has format:
|
|
|
9 #value id type active name path
|
|
|
10 16S_euzby 16S 1 Euzby /usr/local/galaxy/shared/ngs_data/
|
|
|
11 ...
|
|
|
12 """
|
|
|
13 self.reference_bins = []
|
|
|
14
|
|
|
15 if db_spec_path == None: # Default to the command-line lookup table in code's folder:
|
|
|
16 db_spec_path = os.path.join(os.path.dirname(__file__), 'fasta_reference_dbs.tab')
|
|
|
17 self.fieldSpec = common.FieldSpec(db_spec_path)
|
|
|
18
|
|
|
19 def __main__(self): pass
|
|
|
20
|
|
|
21 # Could double check to see if it exists?
|
|
|
22 def build_bins(self, bins, columns):
|
|
|
23
|
|
|
24 if bins == None:
|
|
|
25 self.reference_bins = []
|
|
|
26 return False
|
|
|
27
|
|
|
28 for myfield in bins.strip().strip(';').split(';'):
|
|
|
29 field_spec = myfield.strip().split(':')
|
|
|
30 field_name = field_spec[0].strip()
|
|
|
31
|
|
|
32 if field_name != '':
|
|
|
33 if not field_name.replace('_','').isalnum():
|
|
|
34 common.stop_err("Invalid bin name: " + field_name + ':' + myfield)
|
|
|
35
|
|
|
36 if len(field_spec) < 2: field_spec.append('column') # default grouping = column
|
|
|
37 if len(field_spec) < 3: field_spec.append('') # default filtering = none
|
|
|
38 if len(field_spec) < 4: field_spec.append('') # default no description
|
|
|
39
|
|
|
40 grouping = field_spec[1].strip()
|
|
|
41 if not grouping in ['table', 'column', 'hidden']:
|
|
|
42 common.stop_err("Invalid bin layout: " + grouping)
|
|
|
43
|
|
|
44 bin_filter = field_spec[2].strip()
|
|
|
45 if not bin_filter in ['', 'include', 'exclude']:
|
|
|
46 common.stop_err("Invalid bin sort: " + bin_filter)
|
|
|
47
|
|
|
48 newbin = self.buildBin(field_name, bin_filter)
|
|
|
49 self.reference_bins.append(newbin)
|
|
|
50
|
|
|
51 field = { # any time we have a bin we want sort descending
|
|
|
52 'field': field_name,
|
|
|
53 'group': grouping,
|
|
|
54 'sort': 'desc',
|
|
|
55 'label': newbin.name,
|
|
|
56 'type': 'bin'
|
|
|
57 }
|
|
|
58 columns.append(field)
|
|
|
59 if (field_spec[3] == 'true'): # description field requested
|
|
|
60 field = {
|
|
|
61 'field': field_name + '_desc',
|
|
|
62 'group': 'column',
|
|
|
63 'sort': '', # Allow other sorts????
|
|
|
64 'label': newbin.name + ' Description',
|
|
|
65 'type': 'text'
|
|
|
66 }
|
|
|
67 columns.append(field)
|
|
|
68
|
|
|
69
|
|
|
70 def buildBin(self, bin_folder_name, bin_filter):
|
|
|
71 """
|
|
|
72 Create a lookup table consisting of a dictionary entry for each accession id held in dictionary's file.
|
|
|
73 @param bin_folder_name string name of requested db, e.g 16S_ncbi
|
|
|
74 @param bin_filter string '' or 'include' or 'exclude'
|
|
|
75
|
|
|
76 """
|
|
|
77 bin = ReferenceBin(self.fieldSpec, bin_folder_name, bin_filter)
|
|
|
78
|
|
|
79 try:
|
|
|
80 with open(bin.file_path) as file_in:
|
|
|
81 for line in file_in: # Should always contains succession id
|
|
|
82 #FUTURE: Preprocess so accession ID ready to go.
|
|
|
83 keyValue = line.rstrip().split("\t",1)
|
|
|
84 # keep only first term minus integer portion of id
|
|
|
85 accGeneralId = keyValue[0].split('.')[0]
|
|
|
86 if len(keyValue) >1: description = keyValue[1]
|
|
|
87 else: description = ''
|
|
|
88 bin.lookup[accGeneralId] = description
|
|
|
89
|
|
|
90 file_in.close()
|
|
|
91
|
|
|
92 except IOError:
|
|
|
93 stop_err("Reference bin could not be found or opened: " + self.path + bin_folder_name + '/accession_ids.tab')
|
|
|
94
|
|
|
95 return bin
|
|
|
96
|
|
|
97 def setStatus(self, record):
|
|
|
98
|
|
|
99 if len(self.reference_bins) == 0: return #no bins
|
|
|
100
|
|
|
101 # Use of "extended slices" http://docs.python.org/2.3/whatsnew/section-slices.html
|
|
|
102 # Example sallseqid is 'gi|194753780|ref|XR_046072.1|;gi|195119578|ref|XR_047594.1|;gi|195154052|ref|XR_047967.1|'
|
|
|
103 # Example accs is ['XR_046072.1', 'XR_047594.1', 'XR_047967.1']
|
|
|
104 # Original code was "[1::2][1::2]" (select every 2nd item, then every 2nd item of that)
|
|
|
105 accs = record.sallseqid.split('|')
|
|
|
106
|
|
|
107 if common.re_default_ncbi_id.match(record.sseqid):
|
|
|
108 accs = accs[3::4] #Select every 4th item starting offset 4
|
|
|
109
|
|
|
110 elif common.re_default_ref_id.match(record.sseqid):
|
|
|
111 accs = accs[1::2]
|
|
|
112
|
|
|
113
|
|
|
114 # Check each accession # against each bin.
|
|
|
115 for ptr, bin in enumerate(self.reference_bins):
|
|
|
116 setattr(record, bin.field, '') #Using '','1' not FALSE/TRUE because of tab delim output
|
|
|
117 setattr(record, bin.field + '_desc', '')
|
|
|
118 for acc in accs:
|
|
|
119 accGeneralId = acc.split('.')[0]
|
|
|
120 if accGeneralId in bin.lookup:
|
|
|
121 if bin.exclude: return False
|
|
|
122 setattr(record, bin.field, str(ptr+1))
|
|
|
123 # Include any bin notes for this item
|
|
|
124 setattr(record, bin.field + '_desc', bin.lookup[accGeneralId])
|
|
|
125 break # This result has been binned to this bin so break.
|
|
|
126
|
|
|
127
|
|
|
128
|
|
|
129
|
|
|
130 def __str__(self):
|
|
|
131 return "name: %s dict: %s" % (self.name, str(self.lookup))
|
|
|
132
|
|
|
133 class ReferenceBin:
|
|
|
134 def __init__(self, fieldSpec, bin_folder_name, bin_filter):
|
|
|
135 self.lookup = {}
|
|
|
136 self.folder = bin_folder_name
|
|
|
137 self.name = fieldSpec.getAttribute(bin_folder_name, 'name')
|
|
|
138 self.field = bin_folder_name
|
|
|
139 self.path = fieldSpec.getAttribute(bin_folder_name, 'path')
|
|
|
140 self.exclude = bin_filter
|
|
|
141 #absolute path to reference bins folder: /usr/local/galaxy/shared/ngs_data/
|
|
|
142 self.file_path = os.path.join(self.path + self.folder + '/accession_ids.tab')
|
|
|
143
|
|
|
144 if __name__ == '__main__':
|
|
|
145
|
|
|
146 binManager = ReferenceBins()
|
|
|
147 binManager.__main__()
|
|
|
148
|
