예제 #1
0
    def __call__(self, seq_path=None, result_path=None, log_path=None):
        """Returns a dict mapping {seq_id:(taxonomy, confidence)} for each seq

        Keep in mind, "confidence" is only done for consistency and in fact
        all assignments will have a score of 0 because a method for determining
        confidence is not currently implemented.

        Parameters:
        seq_path: path to file of sequences. The sequences themselves are
            never actually used, but they are needed for their ids.
        result_path: path to file of results. If specified, dumps the
            result to the desired path instead of returning it.
        log_path: path to log, which should include dump of params.
        """

        # initialize the logger
        logger = self._get_logger(log_path)
        logger.info(str(self))

        with open(seq_path, 'U') as f:
            seqs = dict(MinimalFastaParser(f))

        consensus_map = tax2tree.prep_consensus(
            open(self.Params['id_to_taxonomy_fp']),
            seqs.keys())
        seed_con = consensus_map[0].strip().split('\t')[1]
        determine_rank_order(seed_con)

        tipnames_map = load_consensus_map(consensus_map, False)

        tree = load_tree(open(self.Params['tree_fp']), tipnames_map)

        results = tax2tree.generate_constrings(tree, tipnames_map)
        results = tax2tree.clean_output(results, seqs.keys())

        if result_path:
            # if the user provided a result_path, write the
            # results to file
            with open(result_path, 'w') as f:
                for seq_id, (lineage, confidence) in results.iteritems():
                    f.write('%s\t%s\t%s\n' % (seq_id, lineage, confidence))
            logger.info('Result path: %s' % result_path)

        return results
예제 #2
0
    def __call__(self, seq_path=None, result_path=None, log_path=None):
        """Returns a dict mapping {seq_id:(taxonomy, confidence)} for each seq

        Keep in mind, "confidence" is only done for consistency and in fact
        all assignments will have a score of 0 because a method for determining
        confidence is not currently implemented.

        Parameters:
        seq_path: path to file of sequences. The sequences themselves are
            never actually used, but they are needed for their ids.
        result_path: path to file of results. If specified, dumps the
            result to the desired path instead of returning it.
        log_path: path to log, which should include dump of params.
        """

        # initialize the logger
        logger = self._get_logger(log_path)
        logger.info(str(self))

        with open(seq_path, 'U') as f:
            seqs = dict(parse_fasta(f))

        consensus_map = tax2tree.prep_consensus(
            open(self.Params['id_to_taxonomy_fp']),
            seqs.keys())
        seed_con = consensus_map[0].strip().split('\t')[1]
        determine_rank_order(seed_con)

        tipnames_map = load_consensus_map(consensus_map, False)

        tree = load_tree(open(self.Params['tree_fp']), tipnames_map)

        results = tax2tree.generate_constrings(tree, tipnames_map)
        results = tax2tree.clean_output(results, seqs.keys())

        if result_path:
            # if the user provided a result_path, write the
            # results to file
            with open(result_path, 'w') as f:
                for seq_id, (lineage, confidence) in results.iteritems():
                    f.write('%s\t%s\t%s\n' % (seq_id, lineage, confidence))
            logger.info('Result path: %s' % result_path)

        return results