|
kkonganti@101
|
1 #!/usr/bin/env python
|
|
kkonganti@101
|
2
|
|
kkonganti@101
|
3 import os
|
|
kkonganti@101
|
4 import argparse
|
|
kkonganti@101
|
5 import logging as log
|
|
kkonganti@101
|
6 import pandas as pd
|
|
kkonganti@101
|
7 import numpy as np
|
|
kkonganti@101
|
8 from Bio import SeqIO
|
|
kkonganti@101
|
9
|
|
kkonganti@101
|
10
|
|
kkonganti@101
|
11 def main():
|
|
kkonganti@101
|
12 # READ IN ARGUMENTS
|
|
kkonganti@101
|
13 desc = """
|
|
kkonganti@101
|
14 This script is part of the centriflaken pipeline: It processes centrifuge
|
|
kkonganti@101
|
15 output and produces either a filtered FASTQ or a text file of FASTQ IDs based
|
|
kkonganti@101
|
16 on the supplied taxa/bug
|
|
kkonganti@101
|
17 """
|
|
kkonganti@101
|
18 parser = argparse.ArgumentParser(prog='process_centrifuge_output.py', description=desc)
|
|
kkonganti@101
|
19 parser.add_argument("-v", dest='verbose', action="store_true", help="For more verbose output")
|
|
kkonganti@101
|
20 parser.add_argument("-i", dest='input_fastq', required=False,
|
|
kkonganti@101
|
21 help="Path to input FASTQ file (same as input to centrifuge). If not mentioned, \
|
|
kkonganti@101
|
22 a text file of sequence IDs are produced instead of a FASTQ file")
|
|
kkonganti@101
|
23 parser.add_argument("-t", dest='taxa_filtered_fastq_file', required=True,
|
|
kkonganti@101
|
24 help="Path to output FASTQ or output text file filtered by the taxa specified")
|
|
kkonganti@101
|
25 parser.add_argument("-r", dest='cent_report', required=True, help="Path to centrifuge report")
|
|
kkonganti@101
|
26 parser.add_argument("-o", dest='cent_output', required=True, help="Path to centrifuge output")
|
|
kkonganti@101
|
27 parser.add_argument("-b", dest='bug', required=True,
|
|
kkonganti@101
|
28 help="Name or fragment of name of the bug by which reads are extracted")
|
|
kkonganti@101
|
29 args = parser.parse_args()
|
|
kkonganti@101
|
30
|
|
kkonganti@101
|
31 # MORE INFO IF VERBOSE
|
|
kkonganti@101
|
32 if args.verbose:
|
|
kkonganti@101
|
33 log.basicConfig(format="%(levelname)s: %(message)s", level=log.DEBUG)
|
|
kkonganti@101
|
34 else:
|
|
kkonganti@101
|
35 log.basicConfig(format="%(levelname)s: %(message)s")
|
|
kkonganti@101
|
36
|
|
kkonganti@101
|
37 # ASSIGN VARIABLES
|
|
kkonganti@101
|
38 input_fastq = args.input_fastq
|
|
kkonganti@101
|
39 taxa_filtered_fastq_file = args.taxa_filtered_fastq_file
|
|
kkonganti@101
|
40 cent_report = args.cent_report
|
|
kkonganti@101
|
41 cent_output = args.cent_output
|
|
kkonganti@101
|
42 bug = args.bug
|
|
kkonganti@101
|
43 report_col_list = ["name", "taxID"]
|
|
kkonganti@101
|
44 output_col_list = ["taxID", "readID"]
|
|
kkonganti@101
|
45
|
|
kkonganti@101
|
46 # Match and filter taxa names and ids from centrifuge report file
|
|
kkonganti@101
|
47 report_df = pd.read_csv(cent_report, delimiter="\t", usecols=report_col_list)
|
|
kkonganti@101
|
48 report_df['name'] = report_df['name'].str.lower()
|
|
kkonganti@101
|
49 filt_report_df = report_df[report_df['name'].str.contains(bug.lower())]
|
|
kkonganti@101
|
50 #print("\nMatching taxa names and ids:\n",filt_report_df)
|
|
kkonganti@101
|
51 taxID_list = filt_report_df['taxID']
|
|
kkonganti@101
|
52
|
|
kkonganti@101
|
53 # Match the above tax ids to read ids from centrifuge output file and deduplicate
|
|
kkonganti@101
|
54 output_df = pd.read_csv(cent_output, delimiter="\t", usecols=output_col_list)
|
|
kkonganti@101
|
55 filt_output_df = output_df.loc[output_df['taxID'].isin(taxID_list)]
|
|
kkonganti@101
|
56 readID_list = filt_output_df['readID']
|
|
kkonganti@101
|
57 readID_dedup_list = np.unique(readID_list)
|
|
kkonganti@101
|
58 TF=open(taxa_filtered_fastq_file, "w")
|
|
kkonganti@101
|
59
|
|
kkonganti@101
|
60 if (not input_fastq):
|
|
kkonganti@101
|
61 # print("\nFILTERED READ ID LIST:\n", readID_dedup_list)
|
|
kkonganti@101
|
62 for ID in readID_dedup_list:
|
|
kkonganti@101
|
63 TF.write(f"{ID}\n")
|
|
kkonganti@101
|
64 else:
|
|
kkonganti@101
|
65 # Extract filtered reads from input fastq and write to output fastq
|
|
kkonganti@101
|
66 print ("Indexing reads..")
|
|
kkonganti@101
|
67 rec = SeqIO.index(input_fastq,"fastq")
|
|
kkonganti@101
|
68 for i in readID_dedup_list:
|
|
kkonganti@101
|
69 if i in rec:
|
|
kkonganti@101
|
70 SeqIO.write(rec[i], TF, "fastq")
|
|
kkonganti@101
|
71
|
|
kkonganti@101
|
72 TF.close()
|
|
kkonganti@101
|
73
|
|
kkonganti@101
|
74 if __name__ == "__main__":
|
|
kkonganti@101
|
75 main() |
