test parsing blast results

voseq/blast_local/tests.py

@@ -124,3 +124,11 @@ def test_do_blast(self):
         self.blast.save_query_to_file()
         result = self.blast.do_blast()
         self.assertTrue(os.path.isfile(result))
+
+    def test_parse_blast_output(self):
+        self.blast.save_seqs_to_file()
+        self.blast.create_blast_db()
+        self.blast.save_query_to_file()
+        self.blast.do_blast()
+        result = self.blast.parse_blast_output()
+        self.assertTrue(1057 in [i['length'] for i in result])

voseq/blast_local/utils.py

@@ -5,9 +5,10 @@
 import subprocess
 import uuid
 
+from Bio import SeqIO
 from Bio.Blast.Applications import NcbiblastnCommandline
+from Bio.Blast import NCBIXML
 from Bio.Seq import Seq
-from Bio import SeqIO
 from Bio.SeqRecord import SeqRecord
 import pytz
 
@@ -32,6 +33,7 @@ def __init__(self, blast_type, voucher_code, gene_code, mask=None, test=None):
         :param voucher_code:
         :param gene_code:
         """
+        self.e_value = 0.001
         self.blast_type = blast_type
         self.voucher_code = voucher_code
         self.gene_code = gene_code
@@ -184,7 +186,7 @@ def save_query_to_file(self):
 
     def do_blast(self):
         blastn_cline = NcbiblastnCommandline(query=self.query_file, db=self.db,
-                                             evalue=0.001, outfmt=5, out=self.output_file)
+                                             evalue=self.e_value, outfmt=5, out=self.output_file)
         blastn_cline()
         return self.output_file
 
@@ -194,7 +196,20 @@ def parse_blast_output(self):
 
         match_description, max_score, total_score, query_cover, e_value, % ident, accession number
         """
-        pass
+        handle = open(self.output_file, 'r')
+        blast_record = NCBIXML.read(handle)
+        hits = []
+        append = hits.append
+
+        for alignment in blast_record.alignments:
+            for hsp in alignment.hsps:
+                if hsp.expect < self.e_value:
+                    obj = {}
+                    obj['sequence'] = alignment.title
+                    obj['length'] = alignment.length
+                    obj['e_value'] = hsp.expect
+                    append(obj)
+        return hits
 
     def strip_question_marks(self, seq):
         seq = re.sub('^\?+', '', seq)