Fix bug when exporting taxonomy with UNITE database

scripts/format_taxo.py

@@ -6,19 +6,23 @@ if __name__ == "__main__":
     # Load taxo file
     TAXONOMY_FILE = sys.argv[1]
     taxo = pd.read_csv(TAXONOMY_FILE, sep="\t", index_col=0)
-    try :
+    try:
         taxo.drop("#q2:types", inplace=True)
-    except KeyError :
+    except KeyError:
         pass
+
     # Associate taxorank with corresponding prefix letter
-    taxo["Domain"] = ""
+    if taxo.iloc[1]["Taxon"].startswith("k__"):
+        taxo["Kingdom"] = ""  # Unite uses Kingdom rather than Domain
+    else:
+        taxo["Domain"] = ""
     taxo["Phylum"] = ""
     taxo["Class"] = ""
     taxo["Order"] = ""
     taxo["Family"] = ""
     taxo["Genus"] = ""
     taxo["Species"] = ""
-    taxorank_dict = {"d": "Domain", "p": "Phylum", "c": "Class",
+    taxorank_dict = {"k": "Kingdom", "d": "Domain", "p": "Phylum", "c": "Class",
                      "o": "Order", "f": "Family", "g": "Genus", "s": "Species"}
 
     # Create a column for each taxorank
@@ -32,4 +36,4 @@ if __name__ == "__main__":
                 continue
 
     # Export taxo
-    taxo.to_csv(sys.argv[1], sep="\t")
+    taxo.to_csv(sys.argv[1] + ".tsv", sep="\t")