Option to specify the GTF attribute to use for annotaiton.

HISTORY.md

@@ -1,5 +1,7 @@
 - development
 
+  * Add --feature_id as an option to specify the attribute
+    to use in the GTF file for annotation
   * Add gene_id as a 2nd option to add GTF annotation
   * Only do rnafold for precursors shorter than 200nt
 

seqcluster/libs/annotation.py

@@ -6,12 +6,12 @@
 logger = mylog.getLogger("run")
 
 
-def read_gtf_line(cols):
+def read_gtf_line(cols, field="name"):
     """parse gtf line to get class/name information"""
     try:
         group = cols[2]
         attrs = cols[8].split(";")
-        name = [attr.strip().split(" ")[1] for attr in attrs if attr.strip().split(" ")[0].lower().endswith("name")]
+        name = [attr.strip().split(" ")[1] for attr in attrs if attr.strip().split(" ")[0].lower().endswith(field)]
         if not name:
             name = [attr.strip().split(" ")[1] for attr in attrs if attr.strip().split(" ")[0].lower().endswith("gene_id")]
 
@@ -57,7 +57,7 @@ def _position_in_feature(pos_a, pos_b):
     return lento5, lento3, strd
 
 
-def anncluster(c, clus_obj, db, type_ann):
+def anncluster(c, clus_obj, db, type_ann, feature_id="name"):
     """intersect transcription position with annotation files"""
     id_sa, id_ea, id_id, id_idl, id_sta = 1, 2, 3, 4, 5
     if type_ann == "bed":
@@ -72,7 +72,7 @@ def anncluster(c, clus_obj, db, type_ann):
     logger.debug("Type:%s\n" % type_ann)
     for cols in c.features():
         if type_ann == "gtf":
-            cb, sb, eb, stb, db, tag = read_gtf_line(cols[6:])
+            cb, sb, eb, stb, db, tag = read_gtf_line(cols[6:], feature_id)
         else:
             sb = int(cols[id_sb])
             eb = int(cols[id_eb])

seqcluster/libs/parse.py

@@ -179,6 +179,8 @@ def add_subparser_cluster(subparsers):
                        dest="min_seqs", help="threshold to consider a cluster as valid", default=10)
     parser.add_argument("--db",
                         help="prefix for sqlite3 database with results to use htmlViz plugin (in dev).")
+    parser.add_argument("--feature_id",
+                        help="name in GTF to use to annotate clusters", default='name')
     return parser
 
 

seqcluster/make_clusters.py

@@ -224,7 +224,7 @@ def _annotate(args, setclus):
             db = os.path.basename(filebed)
             b = pybedtools.BedTool(filebed)
             c = a.intersect(b, wo=True)
-            setclus = anncluster(c, setclus, db, args.type_ann)
+            setclus = anncluster(c, setclus, db, args.type_ann, args.feature_id)
     return setclus