From 11eb9eece964f8030fff696f8aebfe26117923a9 Mon Sep 17 00:00:00 2001
From: Martijn Vermaat <martijn@vermaat.name>
Date: Wed, 24 Aug 2011 12:01:22 +0000
Subject: [PATCH] Get only primary assembly mappings in position converter
(naive fix for #58).
git-svn-id: https://humgenprojects.lumc.nl/svn/mutalyzer/branches/refactor-mutalyzer-branch@338 eb6bd6ab-9ccd-42b9-aceb-e2899b4a52f1
---
mutalyzer/Db.py | 17 ++++++++----
mutalyzer/Mapper.py | 2 ++
tests/test_converter.py | 54 ++++++++++++++++++++++++++++++++++++
tests/test_variantchecker.py | 5 ----
4 files changed, 68 insertions(+), 10 deletions(-)
create mode 100644 tests/test_converter.py
diff --git a/mutalyzer/Db.py b/mutalyzer/Db.py
index b608cffb..18e8237e 100644
--- a/mutalyzer/Db.py
+++ b/mutalyzer/Db.py
@@ -242,27 +242,34 @@ class Mapping(Db) :
@return: The version number
@rtype: integer
- """
+ @todo: The 'order by chrom asc' is a quick hack to make sure we first
+ get a primary assembly mapping instead of some haplotype mapping
+ for genes in the HLA cluster.
+ A better fix is to return the entire list of mappings, and/or
+ remove all secondary mappings for the HLA cluster.
+ See also test_converter.test_hla_cluster and bug #58.
+ """
q = """
select acc,
txStart, txEnd,
cdsStart, cdsEnd,
exonStarts, exonEnds,
geneName, chrom,
- strand, protAcc,
- MAX(version)
+ strand, protAcc
from map
"""
if version is None:
q += """
- where acc = %s;
+ where acc = %s
+ version desc, order by chrom asc;
"""
statement = (q, mrnaAcc)
else:
q += """
where acc = %s and
- version = %s;
+ version = %s
+ order by chrom asc;
"""
statement = q, (mrnaAcc, version)
diff --git a/mutalyzer/Mapper.py b/mutalyzer/Mapper.py
index 0138e7e2..ef09e108 100644
--- a/mutalyzer/Mapper.py
+++ b/mutalyzer/Mapper.py
@@ -24,6 +24,8 @@ positions to I{g.} notation if the variant is in I{c.} notation or vice versa.
@requires: Modules.Serializers.Transcript
@requires: Bio.Seq.reverse_complement
@requires: collections.defaultdict
+
+@todo: Rename Mapper to converter?
"""
import sys # argv
diff --git a/tests/test_converter.py b/tests/test_converter.py
new file mode 100644
index 00000000..17d4c5c5
--- /dev/null
+++ b/tests/test_converter.py
@@ -0,0 +1,54 @@
+"""
+Tests for the converter (Mapper) module.
+"""
+
+
+#import logging; logging.basicConfig()
+from nose.tools import *
+
+from mutalyzer.config import Config
+from mutalyzer.output import Output
+from mutalyzer.Mapper import Converter
+
+
+class TestConverter():
+ """
+ Test the converter (Mapper) module.
+ """
+ def setUp(self):
+ """
+ Initialize test converter module.
+ """
+ self.config = Config()
+ self.output = Output(__file__, self.config.Output)
+
+ def _converter(self, build):
+ """
+ Create a Converter instance for a given build.
+ """
+ return Converter(build, self.config, self.output)
+
+ def test_converter(self):
+ """
+ Simple test.
+ """
+ converter = self._converter('hg19')
+ genomic = converter.c2chrom('NM_003002.2:c.274G>T')
+ assert_equal(genomic, 'NC_000011.9:g.111959695G>T')
+ coding = converter.chrom2c(genomic, 'list')
+ assert 'NM_003002.2:c.274G>T' in coding
+
+ def test_hla_cluster(self):
+ """
+ Convert to primary assembly.
+
+ Transcript NM_000500.5 is mapped to different chromosome locations,
+ but we like to just see the primary assembly mapping to chromosome 6.
+
+ See also bug #58.
+ """
+ converter = self._converter('hg19')
+ genomic = converter.c2chrom('NM_000500.5:c.92C>T')
+ assert_equal(genomic, 'NC_000006.11:g.32006291C>T')
+ coding = converter.chrom2c(genomic, 'list')
+ assert 'NM_000500.5:c.92C>T' in coding
diff --git a/tests/test_variantchecker.py b/tests/test_variantchecker.py
index 69f9a552..74bbae73 100644
--- a/tests/test_variantchecker.py
+++ b/tests/test_variantchecker.py
@@ -4,13 +4,8 @@ Tests for the variantchecker module.
#import logging; logging.basicConfig()
-import re
-import os
-import random
from nose.tools import *
-from Bio.Seq import Seq
-import mutalyzer
from mutalyzer.config import Config
from mutalyzer.output import Output
from mutalyzer.variantchecker import check_variant
--
GitLab