Mercurial > repos > iuc > humann_unpack_pathways
diff test-data/test-db/metaphlan-db/customizemapping.py @ 2:080ea153677c draft
planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/humann commit 0966faf8782e9043772acfa32f4a4281687a19dd
author | iuc |
---|---|
date | Tue, 07 Feb 2023 19:48:42 +0000 |
parents | |
children |
line wrap: on
line diff
--- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/test-data/test-db/metaphlan-db/customizemapping.py Tue Feb 07 19:48:42 2023 +0000 @@ -0,0 +1,50 @@ +#!/usr/bin/env python +# -*- coding: utf-8 -*- +# Script to generate a extract a custom mapping file from input mapping file. +# Mostly used for a reduced-size demo data generation. + + +import argparse +from pathlib import Path + + +if __name__ == '__main__': + # Read command line + parser = argparse.ArgumentParser(description='Customize HUMAnN utility mapping') + parser.add_argument('--in_mapping', help="Path to mapping file to reduce") + parser.add_argument('--features', help="Path to tabular file with features to keep in first column") + parser.add_argument('--elements', help="Path to tabular file with elements to keep in other columns") + parser.add_argument('--out_mapping', help="Path to reduced mapping file") + args = parser.parse_args() + + in_mapping_fp = Path(args.in_mapping) + feature_fp = Path(args.features) + element_fp = Path(args.elements) + out_mapping_fp = Path(args.out_mapping) + + # extract features to keep + features = set() + with open(feature_fp, 'r') as feature_f: + for line in feature_f.readlines(): + features.add(line.split("\t")[0]) + print(features) + + # extract elements to keep + elements = set() + with open(element_fp, 'r') as element_f: + for line in element_f.readlines(): + elements.add(line.split("\t")[0]) + print(elements) + + # write mapping for features to keep while keeping only elements + with open(in_mapping_fp, 'r') as in_mapping_f: + with open(out_mapping_fp, 'w') as out_mapping_f: + for line in in_mapping_f.readlines(): + l_split = line.split("\t") + feat = l_split[0] + if feat in features: + to_write = [feat] + for e in l_split[1:]: + if e in elements: + to_write.append(e) + out_mapping_f.write("%s\n" % '\t'.join(to_write))