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
|
#!/usr/bin/python
import numpy as np
import os
import sys
from subprocess import *
import platform
import argparse
def main():
"""
You can call this script independently, and will save the
trained matrix as a numpy file.
example: python quikr-train.py -i input.fasta -k 6 -o trained_matrix.npy
"""
parser = argparse.ArgumentParser(description=
" quikr_train returns a custom trained matrix that can be used with \
the quikr function. \n You must supply a kmer. \n ")
parser.add_argument("-i", "--input", help="training database of sequences (fasta format)", required=True)
parser.add_argument("-o", "--output", help="sensing matrix (text file)", required=True)
parser.add_argument("-k", "--kmer", type=int, help="kmer size (integer)", required=False )
args = parser.parse_args()
if not os.path.isfile(args.input):
parser.error( "Input database not found")
# call the quikr train function, save the output with np.save
matrix = quikr_train(args.input, args.kmer)
np.save(args.output, matrix)
return 0
def quikr_train(input_file_location, kmer):
"""
Takes a input fasta file, and kmer, returns a custom trained matrix
"""
kmer_file_name = str(kmer) + "mers.txt"
if not os.path.isfile(kmer_file_name):
print "could not find kmer file"
exit()
uname = platform.uname()[0]
if uname == "Linux":
print "Detected Linux"
input_file = Popen(["./probabilities-by-read-linux", str(kmer), input_file_location, kmer_file_name], stdout=PIPE)
elif uname == "Darwin":
print "Detected Mac OS X"
input_file = Popen(["./probabilities-by-read-osx", str(kmer), input_file_location, kmer_file_name])
# load and normalize the matrix by dividing each element by the sum of it's column.
matrix = np.loadtxt(input_file.stdout)
matrix = np.rot90(matrix)
normalized = matrix / matrix.sum(0)
return normalized
if __name__ == "__main__":
sys.exit(main())
|
