comparison protein_blast_grouping.py @ 1:f2a7dffab581 draft

planemo upload commit 6dde4ec93f27f36a10017393063dcf0568f1d405

0:7abe5f471364

import argparse
import re


class BlastProteinResultParser:
    def __init__(self, blast_file):
        self.blast_file = blast_file

            return len(item[1][key])

        sorted_results = sorted(self.results.items(), key=sort_key, reverse=True)
        return sorted_results[:num_hits]

    def print_results(self, num_hits, sort_key="unique_queries"):
        top_hits = self.get_top_hits(num_hits, sort_key)
        print(f"# Top {num_hits} Hits")
        print(
            "{:<50} {:<25} {:<25}".format(
                "# Name", "Unique Query Matches", "Unique Subject Hits"

        )
        for organism, data in top_hits:
            print(
                "{:<50} {:<25} {:<25}".format(
                    organism, len(data["unique_queries"]), len(data["unique_hits"])
                )
            )


def main():
    parser = argparse.ArgumentParser(

    parser.add_argument("blast", type=argparse.FileType("r"), help="Blast Results")
    parser.add_argument(
        "--hits", type=int, default=5, help="Number of top hits to display"
    )
    parser.add_argument(
        "--sort",
        choices=["unique_queries", "unique_hits"],
        default="unique_queries",
        help="Sort results by 'unique_queries' (default) or 'unique_hits'",
    )
    args = parser.parse_args()

    blast_parser = BlastProteinResultParser(args.blast)
    blast_parser.parse_blast()
    blast_parser.print_results(args.hits, args.sort)


if __name__ == "__main__":
    main()

1:f2a7dffab581

import argparse
import re
import sys


class BlastProteinResultParser:
    def __init__(self, blast_file):
        self.blast_file = blast_file

            return len(item[1][key])

        sorted_results = sorted(self.results.items(), key=sort_key, reverse=True)
        return sorted_results[:num_hits]

    def print_results(
        self, num_hits, sort_key="unique_queries", output_file=sys.stdout
    ):
        top_hits = self.get_top_hits(num_hits, sort_key)
        print(f"# Top {num_hits} Hits")
        print(
            "{:<50} {:<25} {:<25}".format(
                "# Name", "Unique Query Matches", "Unique Subject Hits"

        )
        for organism, data in top_hits:
            print(
                "{:<50} {:<25} {:<25}".format(
                    organism, len(data["unique_queries"]), len(data["unique_hits"])
                ),
                file=output_file,
            )


def main():
    parser = argparse.ArgumentParser(

    parser.add_argument("blast", type=argparse.FileType("r"), help="Blast Results")
    parser.add_argument(
        "--hits", type=int, default=5, help="Number of top hits to display"
    )
    parser.add_argument(
        "--output",
        type=argparse.FileType("w"),
        default="-",
        help="Output file (default: stdout)",
    )
    parser.add_argument(
        "--sort",
        choices=["unique_queries", "unique_hits"],
        default="unique_queries",
        help="Sort results by 'unique_queries' (default) or 'unique_hits'",
    )
    args = parser.parse_args()

    blast_parser = BlastProteinResultParser(args.blast)
    blast_parser.parse_blast()
    blast_parser.print_results(args.hits, args.sort, args.output)

    args.output.close()


if __name__ == "__main__":
    main()