From 9ddcfeacb3191638f42b08af999889d867f0f81c Mon Sep 17 00:00:00 2001
From: Peter Amstutz
Date: Mon, 20 Apr 2020 14:57:25 -0400
Subject: Better handling of duplicate sequences

Also save original fasta label in metadata
---
 workflows/pangenome-generate/merge-metadata.py | 21 ++++++++++++++++++++-
 1 file changed, 20 insertions(+), 1 deletion(-)

(limited to 'workflows/pangenome-generate/merge-metadata.py')

diff --git a/workflows/pangenome-generate/merge-metadata.py b/workflows/pangenome-generate/merge-metadata.py
index 64275b1..bfec781 100644
--- a/workflows/pangenome-generate/merge-metadata.py
+++ b/workflows/pangenome-generate/merge-metadata.py
@@ -1,9 +1,13 @@
+import re
 import schema_salad.schema
 import schema_salad.jsonld_context
+import json
 
 metadataSchema = '$(inputs.metadataSchema.path)'
 metadata = $(inputs.metadata)
 subjects = $(inputs.subjects)
+dups = json.loads('''$(inputs.dups)''')
+originalLabels = $(inputs.originalLabels)
 
 (document_loader,
  avsc_names,
@@ -11,7 +15,22 @@ subjects = $(inputs.subjects)
  metaschema_loader) = schema_salad.schema.load_schema(metadataSchema)
 
 for i, m in enumerate(metadata):
-    doc, metadata = schema_salad.schema.load_and_validate(document_loader, avsc_names, m["path"], True)
+    doc, metadata = schema_salad.schema.load_and_validate(document_loader, avsc_names, m["path"], False, False)
     doc["id"] = subjects[i]
     g = schema_salad.jsonld_context.makerdf(subjects[i], doc, document_loader.ctx)
     print(g.serialize(format="ntriples").decode("utf-8"))
+
+import logging
+
+if dups:
+    sameseqs = open(dups["path"], "rt")
+    for d in sameseqs:
+        logging.warn(d)
+        g = re.match(r"\\d+\\t(.*)", d)
+        logging.warn("%s", g.group(1))
+        sp = g.group(1).split(",")
+        for n in sp[1:]:
+            print("<%s> <http://biohackathon.org/bh20-seq-schema/has_duplicate_sequence> <%s> ." % (n.strip(), sp[0].strip()))
+
+orig = open(originalLabels["path"], "rt")
+print(orig.read())
-- 
cgit v1.2.3