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
|
#!/usr/bin/python
"""
Copyright (c) 2012, Martin S. Lindner,
Robert Koch-Institut, Germany,
All rights reserved.
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are met:
* Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above copyright
notice, this list of conditions and the following disclaimer in the
documentation and/or other materials provided with the distribution.
* The name of the author may not be used to endorse or promote products
derived from this software without specific prior written permission.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND
ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
DISCLAIMED. IN NO EVENT SHALL MARTIN S. LINDNER BE LIABLE FOR ANY
DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
(INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
"""
import numpy as np
import pysam
import sys
import glob
import optparse
from core import gasic
from core import tools
def similarity_correction(names, smat_raw, sam_pattern, bootstrap_samples):
"""
Perform similarity correction step. The similarity matrix and mapping
results must be available.
INPUT:
names: array of genome names
smat_raw: mapping information for similarity matrix with same ordering of genomes as in 'names'
sam_pattern: pattern pointing to the filenames of the SAM-files to analyze
bootstrap_samples: number of bootstrap samples, use 1 to disable bootstrapping
OUTPUT:
total: total number of reads in the dataset
num_reads: number of reads mapped to each genome (array)
corr: abundance of each genome after similarity correction
err: estimated standard error
p: p-value for the confidence, that the true abundance is above some threshold
"""
print("\n--- GASiC correction\n")
# find out the number of reads
total = len( [1 for read in pysam.Samfile(sam_pattern%(names[0]), "r")] )
print(" found %i reads"%total)
# initialize some arrays
# mapping information; mapped[i,j]=1 if read j was successfully mapped to i.
mapped = np.zeros( (len(names), total) )
# total number of successfully mapped reads per reference
num_reads = np.zeros( (len(names),) )
print("\n--- Analyzing SAM files\n")
# analyze the SAM files
for n_ind,nm in enumerate(names):
msg = " SAM file %i of %i"%(n_ind+1,len(names))
sys.stdout.write("\r" + " "*(len(msg)+5) + "\r")
sys.stdout.write(msg)
sys.stdout.flush()
# open SAM file
sf = pysam.Samfile(sam_pattern%nm, "r")
# go through reads in samfile and check if it was successfully mapped
mapped[n_ind,:] = np.array([int(not rd.is_unmapped) for rd in sf])
num_reads[n_ind] = sum(mapped[n_ind,:])
print("\n\n--- Bootstrapping abundances\n")
# run similarity correction step
p,corr,var = gasic.bootstrap(mapped, smat_raw, bootstrap_samples)
err = np.sqrt(var)
print("\nDone estimating abundances\n")
return total,num_reads,corr,err,p
def unique(names, sam_pattern):
""" Determine the number of unique reads for every species based on the read names.
INPUT:
names: array of genome names
sam_pattern: pattern pointing to the filenames of the SAM-files to analyze
OUTPUT:
unique: number of unique reads per species.
"""
# one set for the names of mapped reads for each species
mapped_read_names = [set() for nm in names]
for n,nm in enumerate(names):
# parse the samfile
sf = pysam.Samfile(sam_pattern%nm, "r")
for read in sf:
# add the hash of read name to the set, if read was mapped
if not read.is_unmapped:
mapped_read_names[n].add(hash(read.qname))
unique_read_names = [set() for nm in names]
for n in range(len(names)):
others = set()
for m in range(len(names)):
if n!=m:
others |= mapped_read_names[m]
unique_read_names[n] = mapped_read_names[n] - others
return np.array([len(unq) for unq in unique_read_names])
if __name__=="__main__":
usage = """%prog NAMES
Run the similarity correction step.
Note: Although it is possible to run the read mappers by hand or to create the
similarity matrix manually, we strongly recommend to use the provided Python
scripts 'run_mappers.py' and 'create_similarity_matrix.py'.
Input:
NAMES: Filename of the names file; the plain text names file should
contain one name per line. The name is used as identifier in
the whole algorithm.
See the provided LICENSE file or source code for license information.
"""
# configure the parser
parser = optparse.OptionParser(usage=usage)
parser.add_option('-m', '--similarity-matrix', type='string', dest='smat', default='./similarity_matrix.npy', help='Path to similarity matrix file. The similarity matrix must be created with the same NAMES file. [default: %default]')
parser.add_option('-s', '--samfiles', type='string', dest='sam', default='./SAM/%s.sam', help='Pattern pointing to the SAM files created by the mapper. Placeholder for the name is "%s". [default: %default]')
parser.add_option('-b', '--bootstrap-samples', type='int', dest='boot', default=100, help='Set the number of bootstrap samples. Use 1 to disable bootstrapping [default: %default]')
parser.add_option('-o', '--output', type='string', dest='out', default='./results.txt', help='Plain text output file containing the results. [default: %default]')
# parse arguments
opt, args = parser.parse_args()
numArgs = len(args)
if numArgs == 1:
# read the Names file
names_file = args[0]
names = tools.read_names(names_file)
# load the similarity matrix
smat = np.load(opt.smat)
# start similarity correction
total,num_reads,corr,err,p = similarity_correction(names, smat, opt.sam, opt.boot)
# write results into tab separated file.
ofile = open(opt.out,'w')
ofile.write("#genome name\tmapped reads\testimated reads\testimated error\tp-value\n")
for n_ind,nm in enumerate(names):
# Name, mapped reads, corrected reads, estimated error, p-value
out = "{name}\t{mapped}\t{corr}\t{error}\t{pval}\n"
ofile.write(out.format(name=nm,mapped=num_reads[n_ind],corr=corr[n_ind]*total,error=err[n_ind]*total,pval=p[n_ind]))
ofile.close()
print("--- wrote results to", opt.out)
else:
parser.print_help()
sys.exit(1)
|
