view core.py @ 9:43f6b7f6ebb3 draft

planemo upload commit c50df40caef2fb97c178d6890961e0e527992324-dirty
author cstrittmatter
date Thu, 30 Apr 2020 21:47:42 -0400
parents d0350fe29fdf
children
line wrap: on
line source

#!/usr/bin/env python3


import gzip
import io
import pickle
import os
import sys

from argparse import ArgumentParser
try:
    from .version import SalmID_version
except ImportError:
    SalmID_version = "version unknown"


def reverse_complement(sequence):
    """return the reverse complement of a nucleotide (including IUPAC ambiguous nuceotide codes)"""
    complement = {'A': 'T', 'C': 'G', 'G': 'C', 'T': 'A', 'N': 'N', 'M': 'K', 'R': 'Y', 'W': 'W',
                            'S': 'S', 'Y': 'R', 'K': 'M', 'V': 'B', 'H': 'D', 'D': 'H', 'B': 'V'}
    return "".join(complement[base] for base in reversed(sequence))


def parse_args():
    "Parse the input arguments, use '-h' for help."
    parser = ArgumentParser(description='SalmID - rapid Kmer based Salmonella identifier from sequence data')
    # inputs
    parser.add_argument('-v', '--version', action='version', version='%(prog)s ' + SalmID_version)
    parser.add_argument(
        '-i', '--input_file', type=str, required=False, default='None', metavar='your_fastqgz',
        help='Single fastq.gz file input, include path to file if file is not in same directory ')
    parser.add_argument(
        '-e', '--extension', type=str, required=False, default='.fastq.gz', metavar='file_extension',
        help='File extension, if specified without "--input_dir", SalmID will attempt to ID all files\n' +
             ' with this extension in current directory, otherwise files in input directory')

    parser.add_argument(
        '-d', '--input_dir', type=str, required=False, default='.', metavar='directory',
        help='Directory which contains data for identification, when not specified files in current directory will be analyzed.')
    parser.add_argument(
        '-r', '--report', type=str, required=False, default='percentage', metavar='percentage, coverage or taxonomy',
        help='Report either percentage ("percentage") of clade specific kmers recovered, average kmer-coverage ("cov"), or '
             'taxonomy (taxonomic species ID, plus observed mean k-mer coverages and expected coverage).')
    parser.add_argument(
        '-m', '--mode', type=str, required=False, default='quick', metavar='quick or thorough',
        help='Quick [quick] or thorough [thorough] mode')
    if len(sys.argv) == 1:
        parser.print_help(sys.stderr)
        sys.exit(1)
    return parser.parse_args()


def get_av_read_length(file):
    """Samples the first 100 reads from a fastq file and return the average read length."""
    i = 1
    n_reads = 0
    total_length = 0
    if file.endswith(".gz"):
        file_content = io.BufferedReader(gzip.open(file))
    else:
        file_content = open(file, "r").readlines()
    for line in file_content:
        if i % 4 == 2:
            total_length += len(line.strip())
            n_reads += 1
        i += 1
        if n_reads == 100:
            break
    return total_length / 100


def createKmerDict_reads(list_of_strings, kmer):
    """Count occurence of K-mers in a list of strings

    Args:
        list_of_strings(list of str): nucleotide sequences as a list of strings
        kmer(int): length of the K-mer to count

    Returns:
        dict: dictionary with kmers as keys, counts for each kmer as values"""
    kmer_table = {}
    for string in list_of_strings:
        sequence = string.strip('\n')
        if len(sequence) >= kmer:
            for i in range(len(sequence) - kmer + 1):
                new_mer = sequence[i:i + kmer]
                new_mer_rc = reverse_complement(new_mer)
                if new_mer in kmer_table:
                    kmer_table[new_mer.upper()] += 1
                else:
                    kmer_table[new_mer.upper()] = 1
                if new_mer_rc in kmer_table:
                    kmer_table[new_mer_rc.upper()] += 1
                else:
                    kmer_table[new_mer_rc.upper()] = 1
    return kmer_table


def target_read_kmerizer_multi(file, k, kmerDict_1, kmerDict_2, mode):
    mean_1 = None
    mean_2 = None
    i = 1
    n_reads_1 = 0
    n_reads_2 = 0
    total_coverage_1 = 0
    total_coverage_2 = 0
    reads_1 = []
    reads_2 = []
    total_reads = 0
    if file.endswith(".gz"):
        file_content = io.BufferedReader(gzip.open(file))
    else:
        file_content = open(file, "r").readlines()
    for line in file_content:
        start = int((len(line) - k) // 2)
        if i % 4 == 2:
            total_reads += 1
            if file.endswith(".gz"):
                s1 = line[start:k + start].decode()
                line = line.decode()
            else:
                s1 = line[start:k + start]
            if s1 in kmerDict_1:
                n_reads_1 += 1
                total_coverage_1 += len(line)
                reads_1.append(line)
            if s1 in kmerDict_2:
                n_reads_2 += 1
                total_coverage_2 += len(line)
                reads_2.append(line)
        i += 1
        if mode == 'quick':
            if total_coverage_2 >= 800000:
                break

    if len(reads_1) == 0:
        kmer_Dict1 = {}
    else:
        kmer_Dict1 = createKmerDict_reads(reads_1, k)
        mers_1 = set([key for key in kmer_Dict1])
        mean_1 = sum([kmer_Dict1[key] for key in kmer_Dict1]) / len(mers_1)
    if len(reads_2) == 0:
        kmer_Dict2 = {}
    else:
        kmer_Dict2 = createKmerDict_reads(reads_2, k)
        mers_2 = set([key for key in kmer_Dict2])
        mean_2 = sum([kmer_Dict2[key] for key in kmer_Dict2]) / len(mers_2)
    return kmer_Dict1, kmer_Dict2, mean_1, mean_2, total_reads


def mean_cov_selected_kmers(iterable, kmer_dict, clade_specific_kmers):
    '''
    Given an iterable (list, set, dictrionary) returns mean coverage for the kmers in iterable
    :param iterable: set, list or dictionary containing kmers
    :param kmer_dict: dictionary with kmers as keys, kmer-frequency as value
    :param  clade_specific_kmers: list, dict or set of clade specific kmers
    :return: mean frequency as float
    '''
    if len(iterable) == 0:
        return 0
    return sum([kmer_dict[value] for value in iterable]) / len(clade_specific_kmers)


def kmer_lists(query_fastq_gz, k,
               allmers, allmers_rpoB,
               uniqmers_bongori,
               uniqmers_I,
               uniqmers_IIa,
               uniqmers_IIb,
               uniqmers_IIIa,
               uniqmers_IIIb,
               uniqmers_IV,
               uniqmers_VI,
               uniqmers_VII,
               uniqmers_VIII,
               uniqmers_bongori_rpoB,
               uniqmers_S_enterica_rpoB,
               uniqmers_Escherichia_rpoB,
               uniqmers_Listeria_ss_rpoB,
               uniqmers_Lmono_rpoB,
               mode):
    dict_invA, dict_rpoB, mean_invA, mean_rpoB, total_reads = target_read_kmerizer_multi(query_fastq_gz, k, allmers,
                                                                                         allmers_rpoB, mode)
    target_mers_invA = set([key for key in dict_invA])
    target_mers_rpoB = set([key for key in dict_rpoB])
    if target_mers_invA == 0:
        print('No reads found matching invA, no Salmonella in sample?')
    else:
        p_bongori = (len(uniqmers_bongori & target_mers_invA) / len(uniqmers_bongori)) * 100
        p_I = (len(uniqmers_I & target_mers_invA) / len(uniqmers_I)) * 100
        p_IIa = (len(uniqmers_IIa & target_mers_invA) / len(uniqmers_IIa)) * 100
        p_IIb = (len(uniqmers_IIb & target_mers_invA) / len(uniqmers_IIb)) * 100
        p_IIIa = (len(uniqmers_IIIa & target_mers_invA) / len(uniqmers_IIIa)) * 100
        p_IIIb = (len(uniqmers_IIIb & target_mers_invA) / len(uniqmers_IIIb)) * 100
        p_VI = (len(uniqmers_VI & target_mers_invA) / len(uniqmers_VI)) * 100
        p_IV = (len(uniqmers_IV & target_mers_invA) / len(uniqmers_IV)) * 100
        p_VII = (len(uniqmers_VII & target_mers_invA) / len(uniqmers_VII)) * 100
        p_VIII = (len(uniqmers_VIII & target_mers_invA) / len(uniqmers_VIII)) * 100
        p_bongori_rpoB = (len(uniqmers_bongori_rpoB & target_mers_rpoB) / len(uniqmers_bongori_rpoB)) * 100
        p_Senterica = (len(uniqmers_S_enterica_rpoB & target_mers_rpoB) / len(uniqmers_S_enterica_rpoB)) * 100
        p_Escherichia = (len(uniqmers_Escherichia_rpoB & target_mers_rpoB) / len(uniqmers_Escherichia_rpoB)) * 100
        p_Listeria_ss = (len(uniqmers_Listeria_ss_rpoB & target_mers_rpoB) / len(uniqmers_Listeria_ss_rpoB)) * 100
        p_Lmono = (len(uniqmers_Lmono_rpoB & target_mers_rpoB) / len(uniqmers_Lmono_rpoB)) * 100
        bongori_invA_cov = mean_cov_selected_kmers(uniqmers_bongori & target_mers_invA, dict_invA, uniqmers_bongori)
        I_invA_cov = mean_cov_selected_kmers(uniqmers_I & target_mers_invA, dict_invA, uniqmers_I)
        IIa_invA_cov = mean_cov_selected_kmers(uniqmers_IIa & target_mers_invA, dict_invA, uniqmers_IIa)
        IIb_invA_cov = mean_cov_selected_kmers(uniqmers_IIb & target_mers_invA, dict_invA, uniqmers_IIb)
        IIIa_invA_cov = mean_cov_selected_kmers(uniqmers_IIIa & target_mers_invA, dict_invA, uniqmers_IIIa)
        IIIb_invA_cov = mean_cov_selected_kmers(uniqmers_IIIb & target_mers_invA, dict_invA, uniqmers_IIIb)
        IV_invA_cov = mean_cov_selected_kmers(uniqmers_IV & target_mers_invA, dict_invA, uniqmers_IV)
        VI_invA_cov = mean_cov_selected_kmers(uniqmers_VI & target_mers_invA, dict_invA, uniqmers_VI)
        VII_invA_cov = mean_cov_selected_kmers(uniqmers_VII & target_mers_invA, dict_invA, uniqmers_VII)
        VIII_invA_cov = mean_cov_selected_kmers(uniqmers_VIII & target_mers_invA, dict_invA, uniqmers_VIII)
        S_enterica_rpoB_cov = mean_cov_selected_kmers((uniqmers_S_enterica_rpoB & target_mers_rpoB), dict_rpoB,
                                                      uniqmers_S_enterica_rpoB)
        S_bongori_rpoB_cov = mean_cov_selected_kmers((uniqmers_bongori_rpoB & target_mers_rpoB), dict_rpoB,
                                                     uniqmers_bongori_rpoB)
        Escherichia_rpoB_cov = mean_cov_selected_kmers((uniqmers_Escherichia_rpoB & target_mers_rpoB), dict_rpoB,
                                                       uniqmers_Escherichia_rpoB)
        Listeria_ss_rpoB_cov = mean_cov_selected_kmers((uniqmers_Listeria_ss_rpoB & target_mers_rpoB), dict_rpoB,
                                                       uniqmers_Listeria_ss_rpoB)
        Lmono_rpoB_cov = mean_cov_selected_kmers((uniqmers_Lmono_rpoB & target_mers_rpoB), dict_rpoB,
                                                 uniqmers_Lmono_rpoB)
        coverages = [Listeria_ss_rpoB_cov, Lmono_rpoB_cov, Escherichia_rpoB_cov, S_bongori_rpoB_cov,
                     S_enterica_rpoB_cov, bongori_invA_cov, I_invA_cov, IIa_invA_cov, IIb_invA_cov,
                     IIIa_invA_cov, IIIb_invA_cov, IV_invA_cov, VI_invA_cov, VII_invA_cov, VIII_invA_cov]
        locus_scores = [p_Listeria_ss, p_Lmono, p_Escherichia, p_bongori_rpoB, p_Senterica, p_bongori,
                        p_I, p_IIa, p_IIb, p_IIIa, p_IIIb, p_IV, p_VI, p_VII, p_VIII]
    return locus_scores, coverages, total_reads


def report_taxon(locus_covs, average_read_length, number_of_reads):
    list_taxa = [ 'Listeria ss', 'Listeria monocytogenes', 'Escherichia sp.',  # noqa: E201
                  'Salmonella bongori (rpoB)', 'Salmonella enterica (rpoB)',
                  'Salmonella bongori (invA)', 'S. enterica subsp. enterica (invA)',
                  'S. enterica subsp. salamae (invA: clade a)', 'S. enterica subsp. salamae (invA: clade b)',
                  'S. enterica subsp. arizonae (invA)', 'S. enterica subsp. diarizonae (invA)',
                  'S. enterica subsp. houtenae (invA)', 'S. enterica subsp. indica (invA)',
                  'S. enterica subsp. VII (invA)', 'S. enterica subsp. salamae (invA: clade VIII)' ]  # noqa: E202
    if sum(locus_covs) < 1:
        rpoB = ('No rpoB matches!', 0)
        invA = ('No invA matches!', 0)
        return rpoB, invA, 0.0
    else:
        # given list of scores get taxon
        if sum(locus_covs[0:5]) > 0:
            best_rpoB = max(range(len(locus_covs[1:5])), key=lambda x: locus_covs[1:5][x]) + 1
            all_rpoB = max(range(len(locus_covs[0:5])), key=lambda x: locus_covs[0:5][x])
            if (locus_covs[best_rpoB] != 0) & (all_rpoB == 0):
                rpoB = (list_taxa[best_rpoB], locus_covs[best_rpoB])
            elif (all_rpoB == 0) & (round(sum(locus_covs[1:5]), 1) < 1):
                rpoB = (list_taxa[0], locus_covs[0])
            else:
                rpoB = (list_taxa[best_rpoB], locus_covs[best_rpoB])
        else:
            rpoB = ('No rpoB matches!', 0)
        if sum(locus_covs[5:]) > 0:
            best_invA = max(range(len(locus_covs[5:])), key=lambda x: locus_covs[5:][x]) + 5
            invA = (list_taxa[best_invA], locus_covs[best_invA])
        else:
            invA = ('No invA matches!', 0)
        if 'Listeria' in rpoB[0]:
            return rpoB, invA, (average_read_length * number_of_reads) / 3000000
        else:
            return rpoB, invA, (average_read_length * number_of_reads) / 5000000


def main():
    ex_dir = os.path.dirname(os.path.realpath(__file__))
    args = parse_args()
    input_file = args.input_file
    if input_file != 'None':
        files = [input_file]
    else:
        extension = args.extension
        inputdir = args.input_dir
        files = [inputdir + '/' + f for f in os.listdir(inputdir) if f.endswith(extension)]
    report = args.report
    mode = args.mode
    f_invA = open(ex_dir + "/invA_mers_dict", "rb")
    sets_dict_invA = pickle.load(f_invA)
    f_invA.close()
    allmers = sets_dict_invA['allmers']
    uniqmers_I = sets_dict_invA['uniqmers_I']
    uniqmers_IIa = sets_dict_invA['uniqmers_IIa']
    uniqmers_IIb = sets_dict_invA['uniqmers_IIb']
    uniqmers_IIIa = sets_dict_invA['uniqmers_IIIa']
    uniqmers_IIIb = sets_dict_invA['uniqmers_IIIb']
    uniqmers_IV = sets_dict_invA['uniqmers_IV']
    uniqmers_VI = sets_dict_invA['uniqmers_VI']
    uniqmers_VII = sets_dict_invA['uniqmers_VII']
    uniqmers_VIII = sets_dict_invA['uniqmers_VIII']
    uniqmers_bongori = sets_dict_invA['uniqmers_bongori']

    f = open(ex_dir + "/rpoB_mers_dict", "rb")
    sets_dict = pickle.load(f)
    f.close()

    allmers_rpoB = sets_dict['allmers']
    uniqmers_bongori_rpoB = sets_dict['uniqmers_bongori']
    uniqmers_S_enterica_rpoB = sets_dict['uniqmers_S_enterica']
    uniqmers_Escherichia_rpoB = sets_dict['uniqmers_Escherichia']
    uniqmers_Listeria_ss_rpoB = sets_dict['uniqmers_Listeria_ss']
    uniqmers_Lmono_rpoB = sets_dict['uniqmers_L_mono']
    # todo: run kmer_lists() once, create list of tuples containing data to be used fro different reports
    if report == 'taxonomy':
        print('file\trpoB\tinvA\texpected coverage')
        for f in files:
            locus_scores, coverages, reads = kmer_lists(f, 27,
                                                        allmers, allmers_rpoB,
                                                        uniqmers_bongori,
                                                        uniqmers_I,
                                                        uniqmers_IIa,
                                                        uniqmers_IIb,
                                                        uniqmers_IIIa,
                                                        uniqmers_IIIb,
                                                        uniqmers_IV,
                                                        uniqmers_VI,
                                                        uniqmers_VII,
                                                        uniqmers_VIII,
                                                        uniqmers_bongori_rpoB,
                                                        uniqmers_S_enterica_rpoB,
                                                        uniqmers_Escherichia_rpoB,
                                                        uniqmers_Listeria_ss_rpoB,
                                                        uniqmers_Lmono_rpoB,
                                                        mode)
            pretty_covs = [round(cov, 1) for cov in coverages]
            report = report_taxon(pretty_covs, get_av_read_length(f), reads)
            print(f.split('/')[-1] + '\t' + report[0][0] + '[' + str(report[0][1]) + ']' + '\t' + report[1][0] +
                  '[' + str(report[1][1]) + ']' +
                  '\t' + str(round(report[2], 1)))
    else:
        print(
        'file\tListeria sensu stricto (rpoB)\tL. monocytogenes (rpoB)\tEscherichia spp. (rpoB)\tS. bongori (rpoB)\tS. enterica' +  # noqa: E122
        '(rpoB)\tS. bongori (invA)\tsubsp. I (invA)\tsubsp. II (clade a: invA)\tsubsp. II' +  # noqa: E122
        ' (clade b: invA)\tsubsp. IIIa (invA)\tsubsp. IIIb (invA)\tsubsp.IV (invA)\tsubsp. VI (invA)\tsubsp. VII (invA)' +  # noqa: E122
        '\tsubsp. II (clade VIII : invA)')
        if report == 'percentage':
            for f in files:
                locus_scores, coverages, reads = kmer_lists(f, 27,
                                                            allmers, allmers_rpoB,
                                                            uniqmers_bongori,
                                                            uniqmers_I,
                                                            uniqmers_IIa,
                                                            uniqmers_IIb,
                                                            uniqmers_IIIa,
                                                            uniqmers_IIIb,
                                                            uniqmers_IV,
                                                            uniqmers_VI,
                                                            uniqmers_VII,
                                                            uniqmers_VIII,
                                                            uniqmers_bongori_rpoB,
                                                            uniqmers_S_enterica_rpoB,
                                                            uniqmers_Escherichia_rpoB,
                                                            uniqmers_Listeria_ss_rpoB,
                                                            uniqmers_Lmono_rpoB,
                                                            mode)
                pretty_scores = [str(round(score)) for score in locus_scores]
                print(f.split('/')[-1] + '\t' + '\t'.join(pretty_scores))
        else:
            for f in files:
                locus_scores, coverages, reads = kmer_lists(f, 27,
                                                            allmers, allmers_rpoB,
                                                            uniqmers_bongori,
                                                            uniqmers_I,
                                                            uniqmers_IIa,
                                                            uniqmers_IIb,
                                                            uniqmers_IIIa,
                                                            uniqmers_IIIb,
                                                            uniqmers_IV,
                                                            uniqmers_VI,
                                                            uniqmers_VII,
                                                            uniqmers_VIII,
                                                            uniqmers_bongori_rpoB,
                                                            uniqmers_S_enterica_rpoB,
                                                            uniqmers_Escherichia_rpoB,
                                                            uniqmers_Listeria_ss_rpoB,
                                                            uniqmers_Lmono_rpoB,
                                                            mode)
                pretty_covs = [str(round(cov, 1)) for cov in coverages]
                print(f.split('/')[-1] + '\t' + '\t'.join(pretty_covs))


if __name__ == '__main__':
    main()