Source code

Download source archive ↓

upstream-extended/community/tools/find_subsequences/find_subsequences.py

1#!/usr/bin/env python
2
3import argparse
4import re
5
6from Bio import SeqIO
7from Bio.Alphabet import generic_dna
8from Bio.Seq import Seq
9from Bio.SeqUtils import nt_search
10
11choices = [
12    "embl",
13    "fasta",
14    "fastq-sanger",
15    "fastq",
16    "fastq-solexa",
17    "fastq-illumina",
18    "genbank",
19    "gb",
20]
21
22
23def find_pattern(seqs, pattern, outfile_path, strand):
24    """
25    Finds all occurrences of a pattern in the a given sequence.
26    Outputs sequence ID, start and end postion of the pattern.
27    """
28    pattern = pattern.upper()
29    rev_compl = Seq(pattern, generic_dna).complement()
30    search_func = simple_pattern_search
31    if set(pattern).difference(set("ATCG")):
32        search_func = complex_pattern_search
33
34    with open(outfile_path, "w+") as outfile:
35        for seq in seqs:
36            if strand in ["both", "forward"]:
37                search_func(seq, pattern, outfile)
38            if strand in ["both", "reverse"]:
39                search_func(seq, rev_compl, outfile, "-")
40
41
42def simple_pattern_search(sequence, pattern, outfile, strand="+"):
43    """
44    Simple regular expression search. This is way faster than the complex search.
45    """
46    bed_template = "%s\t%s\t%s\t%s\t%s\t%s\n"
47    for match in re.finditer(str(pattern), str(sequence.seq), re.IGNORECASE):
48        outfile.write(
49            bed_template
50            % (
51                sequence.id,
52                match.start(),
53                match.end(),
54                sequence.description,
55                "",
56                strand,
57            )
58        )
59
60
61def complex_pattern_search(sequence, pattern, outfile, strand="+"):
62    """
63    Searching for pattern with biopyhon's nt_search().
64    This allows for ambiguous values, like N = A or T or C or G, R = A or G ...
65    """
66    pattern_length = len(pattern)
67    matches = nt_search(str(sequence.seq), pattern)
68    bed_template = "%s\t%s\t%s\t%s\t%s\t%s\n"
69    for match in matches[1:]:
70        outfile.write(
71            bed_template
72            % (sequence.id, match, match + pattern_length, sequence.description, "", strand)
73        )
74
75
76if __name__ == "__main__":
77    parser = argparse.ArgumentParser()
78    parser.add_argument("-i", "--input", required=True)
79    parser.add_argument("-o", "--output", required=True)
80    parser.add_argument("-p", "--pattern", required=True)
81    parser.add_argument(
82        "--strand", choices=["both", "forward", "reverse"], default="both"
83    )
84    parser.add_argument("-f", "--format", default="fasta", choices=choices)
85    args = parser.parse_args()
86
87    with open(args.input) as handle:
88        find_pattern(
89            SeqIO.parse(handle, args.format), args.pattern, args.output, args.strand
90        )