diff customizemapping.py @ 0:26ce946f4da9 draft

"planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/humann commit 077b8f34e081e6c427acb0fde0fbb97d1b241e0b"
author iuc
date Wed, 12 May 2021 08:59:30 +0000
parents
children
line wrap: on
line diff
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/customizemapping.py	Wed May 12 08:59:30 2021 +0000
@@ -0,0 +1,47 @@
+#!/usr/bin/env python
+# -*- coding: utf-8 -*-
+
+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))