def test_regeneration(self):
record = create_fake_record()
results = cassis.CassisResults(record.id)
# create a prediction, since it will generate a border with many extra qualifiers
start_marker = ClusterMarker("gene1", Motif(3, 3, score=1))
start_marker.promoter = "gene1"
start_marker.abundance = 2
end_marker = ClusterMarker("gene4", Motif(3, 3, score=1))
end_marker.promoter = "gene3+gene4"
assert end_marker.abundance == 1
cluster = cassis.ClusterPrediction(start_marker, end_marker)
results.subregions = cassis.create_subregions("gene1", [cluster],
record)
assert results.subregions
results.promoters = [
Promoter("gene1", 10, 20, seq=Seq("cgtacgtacgt")),
Promoter("gene2", 30, 40, seq=Seq("cgtacgtacgt")),
CombinedPromoter("gene3", "gene4", 50, 60, seq=Seq("cgtacgtacgt"))
]
round_trip = cassis.regenerate_previous_results(
results.to_json(), record, None)
assert isinstance(round_trip, cassis.CassisResults)
assert len(results.subregions) == len(round_trip.subregions)
for old, new in zip(results.subregions, round_trip.subregions):
assert old.location == new.location
assert old.to_biopython()[0].qualifiers == new.to_biopython(
)[0].qualifiers
assert round_trip.promoters == results.promoters
def test_store_subregions(self):
# this test is similar to test_store_promoters
anchor = "gene3"
start_marker = ClusterMarker("gene1", Motif(3, 3, score=1))
start_marker.promoter = "gene1"
start_marker.abundance = 2
end_marker = ClusterMarker("gene4", Motif(3, 3, score=1))
end_marker.promoter = "gene3+gene4"
assert end_marker.abundance == 1
first_cluster = cassis.ClusterPrediction(start_marker, end_marker)
first_cluster.promoters = 3
first_cluster.genes = 4
start_marker = ClusterMarker("gene1", Motif(4, 4, score=1))
start_marker.promoter = "gene1"
assert start_marker.abundance == 1
end_marker = ClusterMarker("gene5", Motif(4, 4, score=1))
end_marker.promoter = "gene5"
assert end_marker.abundance == 1
second_cluster = cassis.ClusterPrediction(start_marker, end_marker)
second_cluster.promoters = 3
second_cluster.genes = 4
# order reversed because subregions are ordered by length when starts are the same
region_predictions = [second_cluster, first_cluster]
record_with_subregions = create_fake_record()
record_without_subregions = create_fake_record(
) # just the same, without adding subregions
subregions = cassis.create_subregions(anchor, region_predictions,
record_with_subregions)
assert record_with_subregions.get_feature_count(
) == record_without_subregions.get_feature_count()
for region in subregions:
record_with_subregions.add_subregion(region)
# test subregion features
expected_count = record_without_subregions.get_feature_count() + len(
subregions)
assert record_with_subregions.get_feature_count() == expected_count
for i, region in enumerate(region_predictions):
subregion = record_with_subregions.get_subregions()[i]
self.assertEqual(subregion.type, "subregion")
self.assertEqual(subregion.tool, "cassis")
self.assertEqual(subregion.anchor, anchor)
self.assertEqual(subregion.get_qualifier("genes"),
(region.genes, ))
self.assertEqual(subregion.get_qualifier("promoters"),
(region.promoters, ))
self.assertEqual(subregion.get_qualifier("gene_left"),
(region.start.gene, ))
self.assertEqual(subregion.get_qualifier("gene_right"),
(region.end.gene, ))