def test_search_proteins(self):
from views.search import search_proteins
# create 15 genes and proteins
mock_proteins_and_genes(15)
# control: do we start with the mocked proteins not others?
assert not search_proteins('TP53')
# does respect limit? does symbol search work?
results = search_proteins('Gene', 10)
assert results
assert len(results) == 10
assert results[0].name.startswith('Gene')
# are results sorted?
db.session.add_all([
Gene(name=name, preferred_isoform=Protein(refseq='NM_%s' % 20 * i))
for i, name in enumerate(['TPK', 'TPKK'])
])
results = search_proteins('TPK', 2)
assert results[0].name == 'TPK'
assert results[0].best_score < results[1].best_score
# does include both: refseq and symbol search?
assert search_proteins('NM_0003', 1)
# can we change subset of searched features?
assert not search_proteins('NM_0003', 1, features=['gene_symbol'])
assert not search_proteins('Gene', 1, features=['refseq'])
def test_gene_symbol(self):
from search.gene import SymbolGeneSearch
# create 15 genes and proteins
mock_proteins_and_genes(10)
search = SymbolGeneSearch().search
# negative control
assert not search('TP53') # symbol absent
assert not search('NM_0000') # refseq not a symbol
# limiting
results = search('Gene', limit=5)
assert len(results) == 5
assert results[0].name.startswith('Gene')
# should not be case sensitive
results = search('gene')
assert results
# should ignore flanking whitespaces
for query in ('gene ', 'gene ', ' gene', ' gene '):
assert search(query)
def test_protein_name(self):
mock_proteins_and_genes(2)
protein = Protein.query.filter_by(refseq='NM_0001').one()
protein.full_name = 'Important protein'
search = ProteinNameSearch().search
assert search('important')
def test_gene_name(self):
# this is exactly the same as gene_symbol
mock_proteins_and_genes(6)
search = GeneNameSearch().search
results = search('Full name of gene', limit=5)
assert len(results) == 5
assert results[0].full_name.startswith('Full name of gene')
def test_autocomplete_all_proteins(self):
# MC3 GeneList is required as a target (a href for links) where users will be pointed
# after clicking of cancer autocomplete suggestion
gene_list = GeneList(name='TCGA', mutation_source_name=MC3Mutation.name)
db.session.add(gene_list)
mock_proteins_and_genes(15)
response = self.client.get(
'search/autocomplete_all?q=%s' % 'Gene',
follow_redirects=True
)
assert response.status_code == 200
assert response.json['entries'][0]['name'].startswith('Gene')
def test_summary(self):
mock_proteins_and_genes(2)
protein = Protein.query.filter_by(refseq='NM_0001').one()
protein.summary = 'This is an important protein for the FooBar pathway'
search = SummarySearch(minimal_length=3).search
for accepted in ['FooBar', 'foobar', 'foobar pathway']:
assert search(accepted)
# too short
assert not search('an')
# negative control
assert not search('cancer')
def test_refseq(self):
from search.gene import RefseqGeneSearch
# create 15 genes and proteins
mock_proteins_and_genes(10)
search = RefseqGeneSearch().search
# negative control
for phase in ['9999', 'NM_00000', 'Gene']:
assert not search(phase)
# limiting
results = search('NM_', limit=5)
assert len(results) == 5
assert results[0].name.startswith('Gene')
# test the search itself
for refseq in ['NM_0003', 'nm_0003', '0003']:
results = search(refseq)
assert len(results) == 1
assert results[0].name == 'Gene_3'
isoforms = results[0].matched_isoforms
assert len(isoforms) == 1
assert isoforms.pop().refseq == 'NM_0003'
db.session.add_all([
Gene(name='Gene X',
isoforms=[
Protein(refseq='NM_000301'),
Protein(refseq='NM_000302'),
]),
Gene(name='Gene Y', isoforms=[Protein(refseq='NM_000309')])
])
# so there are three genes with isoforms starting with NM_0003
# (those are Gene_3, Gene X, Gene Y). Let see if limiting work
# well when applied per-gene.
queries = {'NM_0003': 2, 'NM_00030': 2, 'NM_000301': 1, 'NM_000302': 1}
for query, expected_result in queries.items():
assert len(search(query, limit=2)) == expected_result
def test_autocomplete_proteins(self):
mock_proteins_and_genes(5)
g = Gene(name='Gene: 6', preferred_isoform=Protein(refseq='NM_006'))
db.session.add(g)
for route in ('autocomplete_proteins', 'proteins'):
for accepted_gene_2_query in ('Gene_2', 'Gene', 'gene', 'Gene_2 ', ' gene', 'gene%20'):
print(route, accepted_gene_2_query)
response = self.client.get(
'search/%s?filters=Search.query:eq:%s' % (route, accepted_gene_2_query),
follow_redirects=True
)
assert response.status_code == 200
assert b'Gene_2' in response.data
assert b'NM_0002' in response.data
response = self.client.get(
'search/autocomplete_proteins?filters=Search.query:eq:Gene: ',
follow_redirects=True
)
assert response.status_code == 200
assert b'Gene: 6' in response.data