-
Notifications
You must be signed in to change notification settings - Fork 89
/
make_kreport.py
executable file
·203 lines (199 loc) · 8.2 KB
/
make_kreport.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
#!/usr/bin/env python
######################################################################
#make_kreport.py takes in the kraken output file and the make_ktaxonomy.py
#output file to generate a kraken report file
#Copyright (C) 2020 Jennifer Lu, [email protected]
#
#This file is part of KrakenTools
#KrakenTools is free software; oyu can redistribute it and/or modify
#it under the terms of the GNU General Public License as published by
#the Free Software Foundation; either version 3 of the license, or
#(at your option) any later version.
#
#This program is distributed in the hope that it will be useful,
#but WITHOUT ANY WARRANTY; without even the implied warranty of
#MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
#GNU General Public License for more details.
#
#You should have received a copy of the GNU General Public License
#along with this program; if not, see <http://www.gnu.org/licenses/>.
#
######################################################################
#Jennifer Lu, [email protected]
#Updated: 04/15/2020
#
#This program creates the kraken report file from
#the make_ktaxonomy.py output and the kraken output file
#
#Required Parameters:
# -i,-k,--kraken X....................kraken output file
# -t,--taxonomy X.....................taxonomy file
# -o, --output X......................output kraken report file
#Optional Parameters:
# -h, --help..........................show help message.
#################################################################################
import os, sys, argparse
import operator
from time import gmtime
from time import strftime
#################################################################################
#Tree Class
#usage: tree node used in constructing taxonomy tree
class Tree(object):
'Tree node.'
def __init__(self, taxid, name, level_rank, level_num, p_taxid, parent=None,children=None):
self.taxid = taxid
self.name = name
self.level_rank= level_rank
self.level_num = int(level_num)
self.p_taxid = p_taxid
self.all_reads = 0
self.lvl_reads = 0
#Parent/children attributes
self.children = []
self.parent = parent
if children is not None:
for child in children:
self.add_child(child)
def add_child(self, node):
assert isinstance(node,Tree)
self.children.append(node)
#################################################################################
#Main method
def main():
#Parse arguments
parser = argparse.ArgumentParser()
parser.add_argument('-i','--input', '-k','--kraken', dest='kraken_file', required=True,
help='Kraken output file (5 tab-delimited columns, taxid in 3rd column)')
parser.add_argument('-t','--taxonomy', dest='tax_file', required=True,
help='Output taxonomy file from make_ktaxonomy.py')
parser.add_argument('-o','--output',dest='out_file', required=True,
help='Output kraken report file')
parser.add_argument('--use-read-len',dest='use_read_len',
action='store_true',default=False, required=False,
help='Make report file using sum of read lengths [default: read counts]')
args = parser.parse_args()
#Start Program
time = strftime("%m-%d-%Y %H:%M:%S", gmtime())
sys.stdout.write("PROGRAM START TIME: " + time + '\n')
#STEP 1/4: READ TAXONOMY FILE
count_nodes = 0
sys.stdout.write(">> STEP 1/4: Reading taxonomy %s...\n" % args.tax_file)
sys.stdout.write("\t%i nodes saved" % (count_nodes))
#Parse taxonomy file
root_node = -1
taxid2node = {}
t_file = open(args.tax_file,'r')
for line in t_file:
count_nodes += 1
sys.stdout.write("\r\t%i nodes saved" % (count_nodes))
sys.stdout.flush()
[taxid, p_tid, rank, lvl_num, name] = line.strip().split('\t|\t')
curr_node = Tree(taxid, name, rank, lvl_num, p_tid)
taxid2node[taxid] = curr_node
#set parent/kids
if taxid == "1":
root_node = curr_node
else:
curr_node.parent = taxid2node[p_tid]
taxid2node[p_tid].add_child(curr_node)
t_file.close()
sys.stdout.write("\r\t%i nodes saved\n" % (count_nodes))
sys.stdout.flush()
#STEP 2/4: READ KRAKEN FILE FOR COUNTS PER TAXID
read_count = 0
sys.stdout.write(">> STEP 2/4: Reading kraken file %s...\n" % args.kraken_file)
sys.stdout.write("\t%i million reads processed" % read_count)
sys.stdout.flush()
#Save counts per taxid
taxid2counts = {}
taxid2allcounts = {}
k_file = open(args.kraken_file,'r')
for line in k_file:
read_count += 1
if read_count % 1000 == 0:
sys.stdout.write('\r\t%0.3f million reads processed' % float(read_count/1000000.))
sys.stdout.flush()
l_vals = line.strip().split('\t')
taxid = l_vals[2]
count = 1
#If using read length instead of read counts
if args.use_read_len:
if '|' in l_vals[3]:
[len1,len2] = l_vals[3].split('|')
count = int(len1)+int(len2)
else:
count = int(l_vals[3])
#add to dictionaries
if taxid not in taxid2counts:
taxid2counts[taxid] = count
taxid2allcounts[taxid] = count
else:
taxid2counts[taxid] += count
taxid2allcounts[taxid] += count
k_file.close()
sys.stdout.write('\r\t%0.3f million reads processed\n' % float(read_count/1000000.))
sys.stdout.flush()
#STEP 3/4: FOR EVERY TAXID PARSED, ADD UP TOTAL READS
sys.stdout.write(">> STEP 3/4: Creating final tree...\n")
for curr_tid in taxid2counts:
#Skip unclassified
if curr_tid == '0':
continue
p_node = taxid2node[curr_tid].parent
add_counts = taxid2counts[curr_tid]
#Assign reads for node
taxid2node[curr_tid].lvl_reads += add_counts
taxid2node[curr_tid].all_reads += add_counts
while (p_node != None):
#Add child reads to parent node
p_taxid = p_node.taxid
if p_taxid not in taxid2allcounts:
taxid2allcounts[p_taxid] = add_counts
else:
taxid2allcounts[p_taxid] += add_counts
p_node.all_reads += add_counts
#Get next parent node
p_node = p_node.parent
#STEP 4/4: PRINT REPORT FILE
sys.stdout.write(">> STEP 4/4: Printing report file to %s...\n" % args.out_file)
o_file = open(args.out_file,'w')
#Write line for unclassified reads:
if '0' in taxid2counts:
o_file.write("%6.2f\t" % (float(taxid2counts['0'])/float(read_count)*100))
o_file.write("%i\t%i\t" % (taxid2counts['0'],taxid2counts['0']))
o_file.write('U\t0\tunclassified\n')
#Get remaining lines
parse_nodes = [root_node]
while len(parse_nodes) > 0:
curr_node = parse_nodes.pop(0)
curr_tid = curr_node.taxid
#Print information for this level
o_file.write("%6.2f\t" % (float(taxid2allcounts[curr_tid])/float(read_count)*100))
o_file.write("%i\t" % taxid2allcounts[curr_tid])
if curr_tid not in taxid2counts:
o_file.write("0\t")
else:
o_file.write("%i\t" % taxid2counts[curr_tid])
o_file.write("%s\t" % curr_node.level_rank)
o_file.write("%s\t" % curr_tid)
o_file.write(" "*curr_node.level_num*2 + curr_node.name + "\n")
#Add children to list
for child in sorted(curr_node.children, key=operator.attrgetter('all_reads')):
if child.taxid not in taxid2allcounts:
continue
if taxid2allcounts[child.taxid] == 0:
continue
#Add to list
parse_nodes.insert(0,child)
o_file.close()
#End of program
time = strftime("%m-%d-%Y %H:%M:%S", gmtime())
sys.stdout.write("PROGRAM END TIME: " + time + '\n')
sys.exit(0)
#################################################################################
if __name__ == "__main__":
main()
#################################################################################
##################################END OF PROGRAM#################################
#################################################################################