def test_NIC_match(self):
""" Example where the transcript is an NIC match to an existing one by
virtue of skipping an exon.
"""
conn, cursor = get_db_cursor()
build = "toy_build"
edge_dict = talon.make_edge_dict(cursor)
location_dict = talon.make_location_dict(build, cursor)
run_info = talon.init_run_info(cursor, build)
transcript_dict = talon.make_transcript_dict(cursor, build)
vertex_2_gene = talon.make_vertex_2_gene_dict(cursor)
gene_starts, gene_ends = talon.make_gene_start_and_end_dict(
cursor, build)
chrom = "chr1"
positions = [1, 100, 900, 1000]
edge_IDs = [run_info.edge + 1]
vertex_IDs = [2, 5]
strand = "+"
v_novelty = [0, 0]
gene_ID, transcript_ID, novelty, start_end_info = talon.process_NIC(
chrom, positions, strand, edge_IDs, vertex_IDs, transcript_dict,
gene_starts, gene_ends, edge_dict, location_dict, vertex_2_gene,
run_info)
correct_gene_ID = fetch_correct_ID("TG1", "gene", cursor)
assert gene_ID == correct_gene_ID
assert start_end_info["vertex_IDs"] == [1, 2, 5, 6]
assert transcript_dict[frozenset(start_end_info["edge_IDs"])] != None
conn.close()
def test_partial_match_3prime(self):
""" Example where the transcript is short, so it overlaps the
annotated transcript but is not an accepted match.
the end should get assigned to the annotated end, but the end is
novel """
conn, cursor = get_db_cursor()
build = "toy_build"
talon.make_temp_novel_gene_table(cursor, build)
talon.make_temp_monoexonic_transcript_table(cursor, build)
edge_dict = talon.make_edge_dict(cursor)
location_dict = talon.make_location_dict(build, cursor)
run_info = talon.init_run_info(cursor, build)
transcript_dict = talon.make_transcript_dict(cursor, build)
vertex_2_gene = talon.make_vertex_2_gene_dict(cursor)
gene_starts, gene_ends = talon.make_gene_start_and_end_dict(cursor, build)
chrom = "chr4"
strand = "-"
positions = ( 2000, 1100 )
annotation = talon.identify_monoexon_transcript(chrom, positions,
strand, cursor,
location_dict, edge_dict,
transcript_dict, vertex_2_gene,
gene_starts, gene_ends, run_info)
correct_gene_ID = fetch_correct_ID("TG6", "gene", cursor)
assert annotation['gene_ID'] == correct_gene_ID
assert annotation['start_delta'] == None
assert annotation['end_delta'] == -100
conn.close()
def test_NNC(self):
""" Example where the transcript skips an exon and has a novel splice
donor
"""
conn, cursor = get_db_cursor()
build = "toy_build"
talon.make_temp_novel_gene_table(cursor, build)
edge_dict = talon.make_edge_dict(cursor)
location_dict = talon.make_location_dict(build, cursor)
run_info = talon.init_run_info(cursor, build)
transcript_dict = talon.make_transcript_dict(cursor, build)
vertex_2_gene = talon.make_vertex_2_gene_dict(cursor)
gene_starts, gene_ends = talon.make_gene_start_and_end_dict(
cursor, build)
chrom = "chr1"
strand = "+"
positions = [1, 50, 900, 1000]
annotation = talon.identify_transcript(chrom, positions, strand,
cursor, location_dict,
edge_dict, transcript_dict,
vertex_2_gene, gene_starts,
gene_ends, run_info)
correct_gene_ID = fetch_correct_ID("TG1", "gene", cursor)
novelty_types = [x[-2] for x in annotation['transcript_novelty']]
assert annotation['gene_ID'] == correct_gene_ID
assert "NNC_transcript" in novelty_types
assert annotation['start_delta'] == annotation['end_delta'] == 0
conn.close()
def test_ISM_internal(self):
""" Example where the transcript matches an internal exon
"""
conn, cursor = get_db_cursor()
build = "toy_build"
talon.make_temp_novel_gene_table(cursor, build)
talon.make_temp_monoexonic_transcript_table(cursor, build)
edge_dict = talon.make_edge_dict(cursor)
location_dict = talon.make_location_dict(build, cursor)
run_info = talon.init_run_info(cursor, build)
transcript_dict = talon.make_transcript_dict(cursor, build)
vertex_2_gene = talon.make_vertex_2_gene_dict(cursor)
gene_starts, gene_ends = talon.make_gene_start_and_end_dict(
cursor, build)
chrom = "chr1"
strand = "+"
positions = (500, 600)
annotation = talon.identify_monoexon_transcript(
chrom, positions, strand, cursor, location_dict, edge_dict,
transcript_dict, vertex_2_gene, gene_starts, gene_ends, run_info)
correct_gene_ID = fetch_correct_ID("TG1", "gene", cursor)
novelty_types = [x[-2] for x in annotation['transcript_novelty']]
assert annotation['gene_ID'] == correct_gene_ID
assert "ISM_transcript" in novelty_types
assert annotation['start_delta'] == annotation['end_delta'] == 0
conn.close()
def test_FSM_perfect(self):
""" Example where the transcript is a perfect full splice match.
"""
conn, cursor = get_db_cursor()
build = "toy_build"
talon.make_temp_novel_gene_table(cursor, build)
edge_dict = talon.make_edge_dict(cursor)
location_dict = talon.make_location_dict(build, cursor)
run_info = talon.init_run_info(cursor, build)
transcript_dict = talon.make_transcript_dict(cursor, build)
vertex_2_gene = talon.make_vertex_2_gene_dict(cursor)
gene_starts, gene_ends = talon.make_gene_start_and_end_dict(
cursor, build)
chrom = "chr1"
strand = "+"
positions = [1, 100, 500, 600, 900, 1000]
annotation = talon.identify_transcript(chrom, positions, strand,
cursor, location_dict,
edge_dict, transcript_dict,
vertex_2_gene, gene_starts,
gene_ends, run_info)
correct_gene_ID = fetch_correct_ID("TG1", "gene", cursor)
correct_transcript_ID = fetch_correct_ID("TG1-001", "transcript",
cursor)
assert annotation['gene_ID'] == correct_gene_ID
assert annotation['transcript_ID'] == correct_transcript_ID
assert annotation['transcript_novelty'] == []
conn.close()
def test_FSM_end_diff(self):
""" Example where the transcript is an FSM but has a difference on
the ends large enough to be novel.
"""
conn, cursor = get_db_cursor()
build = "toy_build"
edge_dict = talon.make_edge_dict(cursor)
location_dict = talon.make_location_dict(build, cursor)
run_info = talon.init_run_info(cursor, build)
transcript_dict = talon.make_transcript_dict(cursor, build)
orig_vertices = run_info['vertex']
gene_starts, gene_ends = talon.make_gene_start_and_end_dict(cursor, build)
chrom = "chr2"
positions = [1, 100, 500, 600, 900, 1301] #Last postion is > 300bp away
strand = "+"
edge_IDs = [13, 14, 15]
vertex_IDs = [14, 15, 16, 17]
v_novelty = [0, 0, 0, 0]
all_matches = talon.search_for_ISM(edge_IDs, transcript_dict)
gene_ID, transcript_ID, novelty, start_end_info = talon.process_FSM(chrom,
positions, strand, edge_IDs,
vertex_IDs, all_matches,
gene_starts, gene_ends,
edge_dict,
location_dict, run_info)
correct_gene_ID = fetch_correct_ID("TG2", "gene", cursor)
correct_transcript_ID = fetch_correct_ID("TG2-001", "transcript", cursor)
assert gene_ID == correct_gene_ID
assert transcript_ID == correct_transcript_ID
assert start_end_info["end_vertex"] == orig_vertices + 1
conn.close()
def test_genomic_unspliced(self):
""" Monoexonic fragment that overlaps gene 1 """
conn, cursor = get_db_cursor()
build = "toy_build"
talon.make_temp_novel_gene_table(cursor, build)
talon.make_temp_monoexonic_transcript_table(cursor, build)
edge_dict = talon.make_edge_dict(cursor)
location_dict = talon.make_location_dict(build, cursor)
run_info = talon.init_run_info(cursor, build)
transcript_dict = talon.make_transcript_dict(cursor, build)
vertex_2_gene = talon.make_vertex_2_gene_dict(cursor)
gene_starts, gene_ends = talon.make_gene_start_and_end_dict(
cursor, build)
chrom = "chr1"
strand = "+"
positions = (1, 990)
annotation = talon.identify_monoexon_transcript(
chrom, positions, strand, cursor, location_dict, edge_dict,
transcript_dict, vertex_2_gene, gene_starts, gene_ends, run_info)
correct_gene_ID = fetch_correct_ID("TG1", "gene", cursor)
novelty_types = [x[-2] for x in annotation['transcript_novelty']]
assert annotation['gene_ID'] == correct_gene_ID
assert "genomic_transcript" in novelty_types
assert annotation['end_delta'] == -10
conn.close()
def test_FSM_end_diff(self):
""" Example where the transcript is an FSM but has a difference on
the ends large enough to be novel.
"""
conn, cursor = get_db_cursor()
build = "toy_build"
talon.make_temp_novel_gene_table(cursor, build)
edge_dict = talon.make_edge_dict(cursor)
location_dict = talon.make_location_dict(build, cursor)
run_info = talon.init_run_info(cursor, build)
transcript_dict = talon.make_transcript_dict(cursor, build)
vertex_2_gene = talon.make_vertex_2_gene_dict(cursor)
gene_starts, gene_ends = talon.make_gene_start_and_end_dict(
cursor, build)
chrom = "chr2"
strand = "+"
positions = [1, 100, 500, 600, 900, 1500]
annotation = talon.identify_transcript(chrom, positions, strand,
cursor, location_dict,
edge_dict, transcript_dict,
vertex_2_gene, gene_starts,
gene_ends, run_info)
correct_gene_ID = fetch_correct_ID("TG2", "gene", cursor)
novelty_types = [x[-2] for x in annotation['transcript_novelty']]
assert annotation['gene_ID'] == correct_gene_ID
assert annotation['end_delta'] == None
conn.close()
def test_no_match(self):
""" Example with no FSM match """
conn, cursor = get_db_cursor()
build = "toy_build"
edge_dict = talon.make_edge_dict(cursor)
location_dict = talon.make_location_dict(build, cursor)
run_info = talon.init_run_info(cursor, build)
transcript_dict = talon.make_transcript_dict(cursor, build)
gene_starts, gene_ends = talon.make_gene_start_and_end_dict(cursor, build)
chrom = "chr1"
positions = [1, 100, 500, 600]
strand = "+"
edge_IDs = [2]
vertex_IDs = [2,3,4,5]
v_novelty = [0, 0, 0, 0]
all_matches = talon.search_for_ISM(edge_IDs, transcript_dict)
gene_ID, transcript_ID, novelty, start_end_info = talon.process_FSM(chrom,
positions, strand, edge_IDs,
vertex_IDs, all_matches,
gene_starts, gene_ends,
edge_dict,
location_dict, run_info)
assert gene_ID == transcript_ID == None
conn.close()
def test_ISM_suffix(self):
""" Example where the transcript is an ISM with suffix
"""
conn, cursor = get_db_cursor()
build = "toy_build"
edge_dict = talon.make_edge_dict(cursor)
location_dict = talon.make_location_dict(build, cursor)
run_info = talon.init_run_info(cursor, build)
transcript_dict = talon.make_transcript_dict(cursor, build)
gene_starts, gene_ends = talon.make_gene_start_and_end_dict(
cursor, build)
chrom = "chr1"
strand = "+"
positions = [500, 600, 900, 1000]
edge_IDs = [4]
vertex_IDs = [4, 5]
v_novelty = [0, 0]
all_matches = talon.search_for_ISM(edge_IDs, transcript_dict)
gene_ID, transcript_ID, novelty, start_end_info = talon.process_ISM(
chrom, positions, strand, edge_IDs, vertex_IDs, all_matches,
transcript_dict, gene_starts, gene_ends, edge_dict, location_dict,
run_info)
correct_gene_ID = fetch_correct_ID("TG1", "gene", cursor)
assert gene_ID == correct_gene_ID
assert start_end_info["vertex_IDs"] == [3, 4, 5, 6]
assert start_end_info["edge_IDs"] == [3, 4, 5]
assert start_end_info["start_novelty"] == 0 # because the exon is known
assert start_end_info["end_novelty"] == 0
assert transcript_dict[frozenset(start_end_info["edge_IDs"])] != None
conn.close()
def test_ISM_prefix(self):
""" Example where the transcript is a prefix ISM with a novel start
"""
conn, cursor = get_db_cursor()
build = "toy_build"
edge_dict = talon.make_edge_dict(cursor)
location_dict = talon.make_location_dict(build, cursor)
run_info = talon.init_run_info(cursor, build)
transcript_dict = talon.make_transcript_dict(cursor, build)
gene_starts, gene_ends = talon.make_gene_start_and_end_dict(
cursor, build)
orig_exons = run_info["edge"]
chrom = "chr1"
strand = "+"
positions = [1, 100, 500, 600]
edge_IDs = [2]
vertex_IDs = [2, 3]
v_novelty = [0, 0]
all_matches = talon.search_for_ISM(edge_IDs, transcript_dict)
gene_ID, transcript_ID, novelty, start_end_info = talon.process_ISM(
chrom, positions, strand, edge_IDs, vertex_IDs, all_matches,
transcript_dict, gene_starts, gene_ends, edge_dict, location_dict,
run_info)
correct_gene_ID = fetch_correct_ID("TG1", "gene", cursor)
assert gene_ID == correct_gene_ID
assert start_end_info["vertex_IDs"] == [1, 2, 3, 4]
assert start_end_info["edge_IDs"] == [1, 2, 3]
conn.close()
def test_antisense(self):
""" Example where the transcript is antisense """
conn, cursor = get_db_cursor()
build = "toy_build"
talon.make_temp_novel_gene_table(cursor, build)
talon.make_temp_monoexonic_transcript_table(cursor, build)
edge_dict = talon.make_edge_dict(cursor)
location_dict = talon.make_location_dict(build, cursor)
run_info = talon.init_run_info(cursor, build)
transcript_dict = talon.make_transcript_dict(cursor, build)
vertex_2_gene = talon.make_vertex_2_gene_dict(cursor)
gene_starts, gene_ends = talon.make_gene_start_and_end_dict(cursor, build)
chrom = "chr4"
strand = "+"
positions = ( 1300, 3900 )
annotation = talon.identify_monoexon_transcript(chrom, positions,
strand, cursor,
location_dict, edge_dict,
transcript_dict, vertex_2_gene,
gene_starts, gene_ends, run_info)
anti_gene_ID = fetch_correct_ID("TG6", "gene", cursor)
gene_novelty_types = [ x[-2] for x in annotation['gene_novelty']]
t_novelty_types = [ x[-2] for x in annotation['transcript_novelty']]
assert annotation['gene_novelty'][0][-1] == "TRUE"
assert "antisense_gene" in gene_novelty_types
assert "antisense_transcript" in t_novelty_types
conn.close()
def test_spliced_antisense(self):
""" Example where the transcript matches known vertices but is antisense
"""
conn, cursor = get_db_cursor()
build = "toy_build"
talon.make_temp_novel_gene_table(cursor, build)
edge_dict = talon.make_edge_dict(cursor)
location_dict = talon.make_location_dict(build, cursor)
run_info = talon.init_run_info(cursor, build)
transcript_dict = talon.make_transcript_dict(cursor, build)
vertex_2_gene = talon.make_vertex_2_gene_dict(cursor)
gene_starts, gene_ends = talon.make_gene_start_and_end_dict(
cursor, build)
chrom = "chr2"
strand = "-"
positions = [1000, 900, 600, 500, 100, 1]
annotation = talon.identify_transcript(chrom, positions, strand,
cursor, location_dict,
edge_dict, transcript_dict,
vertex_2_gene, gene_starts,
gene_ends, run_info)
anti_gene_ID = fetch_correct_ID("TG2", "gene", cursor)
gene_novelty_types = [x[-2] for x in annotation['gene_novelty']]
t_novelty_types = [x[-2] for x in annotation['transcript_novelty']]
assert annotation['gene_novelty'][0][-1] == "TRUE"
assert "antisense_gene" in gene_novelty_types
assert "antisense_transcript" in t_novelty_types
assert annotation['start_delta'] == annotation['end_delta'] == 0
conn.close()
def test_NIC_instead_of_ISM(self):
""" Test case where the transcript looks like an ISM, but is NIC on
account of having known starts and ends """
conn, cursor = get_db_cursor()
build = "toy_build"
talon.make_temp_novel_gene_table(cursor, build)
edge_dict = talon.make_edge_dict(cursor)
location_dict = talon.make_location_dict(build, cursor)
run_info = talon.init_run_info(cursor, build)
transcript_dict = talon.make_transcript_dict(cursor, build)
vertex_2_gene = talon.make_vertex_2_gene_dict(cursor)
gene_starts, gene_ends = talon.make_gene_start_and_end_dict(
cursor, build)
chrom = "chr3"
strand = "+"
positions = (800, 1000, 1200, 1400, 1600, 1800, 2000, 2200)
annotation = talon.identify_transcript(chrom, positions, strand,
cursor, location_dict,
edge_dict, transcript_dict,
vertex_2_gene, gene_starts,
gene_ends, run_info)
correct_gene_ID = fetch_correct_ID("TG5", "gene", cursor)
novelty_types = [x[-2] for x in annotation['transcript_novelty']]
assert annotation['gene_ID'] == correct_gene_ID
assert "NIC_transcript" in novelty_types
conn.close()
def test_match(self):
""" Example where the transcript is a moniexonic match.
"""
conn, cursor = get_db_cursor()
build = "toy_build"
talon.make_temp_novel_gene_table(cursor, build)
talon.make_temp_monoexonic_transcript_table(cursor, build)
edge_dict = talon.make_edge_dict(cursor)
location_dict = talon.make_location_dict(build, cursor)
run_info = talon.init_run_info(cursor, build)
transcript_dict = talon.make_transcript_dict(cursor, build)
vertex_2_gene = talon.make_vertex_2_gene_dict(cursor)
gene_starts, gene_ends = talon.make_gene_start_and_end_dict(cursor, build)
chrom = "chr4"
strand = "-"
positions = ( 3900, 1100 )
annotation = talon.identify_monoexon_transcript(chrom, positions,
strand, cursor,
location_dict, edge_dict,
transcript_dict, vertex_2_gene,
gene_starts, gene_ends, run_info)
correct_gene_ID = fetch_correct_ID("TG6", "gene", cursor)
correct_transcript_ID = fetch_correct_ID("TG6-001", "transcript", cursor)
assert annotation['gene_ID'] == correct_gene_ID
assert annotation['start_delta'] == 100
assert annotation['end_delta'] == -100
conn.close()
def test_find_match(self):
""" Example where the toy transcript edge dict does not contain the
edge being queried.
"""
conn, cursor = get_db_cursor()
# Create a location dict and then fetch vertices for two psotions
build = "toy_build"
location_dict = talon.make_location_dict(build, cursor)
edge_dict = talon.make_edge_dict(cursor)
conn.close()
chrom = "chr1"
pos1 = 600
pos2 = 500
v1 = talon.search_for_vertex_at_pos(chrom, pos1,
location_dict)["location_ID"]
v2 = talon.search_for_vertex_at_pos(chrom, pos2,
location_dict)["location_ID"]
assert v1 != None
assert v2 != None
# Now look for the edge between them
edge_match = talon.search_for_edge(v1, v2, "exon", edge_dict)
assert edge_match == None
# Try them in the opposite order
edge_match = talon.search_for_edge(v2, v1, "exon", edge_dict)
assert edge_match["edge_ID"] == 3
def test_antisense(self):
""" Example where all of the vertices are in the database, but the edges
are not, because they are antisense to the original transcript """
conn, cursor = get_db_cursor()
build = "toy_build"
edge_dict = talon.make_edge_dict(cursor)
run_info = talon.init_run_info(cursor, build)
orig_n_edges = len(edge_dict)
conn.close()
chrom = "chr2"
vertex_IDs = [14, 13, 12, 11, 10, 9]
strand = "-"
edge_IDs, novelty = talon.match_all_transcript_edges(
vertex_IDs, strand, edge_dict, run_info)
expected_edges = []
for i in range(1, 6):
num = orig_n_edges + i
edge_id = num
expected_edges.append(edge_id)
assert edge_IDs == tuple(expected_edges)
assert novelty == (1, 1, 1, 1, 1)
def test_datasets(self):
""" Try to add dataset metadata to database """
conn, cursor = get_db_cursor()
build = "toy_build"
edge_dict = talon.make_edge_dict(cursor)
run_info = talon.init_run_info(cursor, build)
datasets = [(1, "toy", "toy", "toy")]
talon.add_datasets(cursor, datasets)
# Test if items are there
query = "SELECT * FROM dataset"
cursor.execute(query)
assert len(cursor.fetchall()) == 1
conn.close()
def test_antisense(self):
""" Example where the vertices are known but there is no same-strand
match """
conn, cursor = get_db_cursor()
build = "toy_build"
edge_dict = talon.make_edge_dict(cursor)
locations = talon.make_location_dict(build, cursor)
run_info = talon.init_run_info(cursor, build)
transcript_dict = talon.make_transcript_dict(cursor, build)
vertex_2_gene = talon.make_vertex_2_gene_dict(cursor)
gene_starts, gene_ends = talon.make_gene_start_and_end_dict(
cursor, build)
# Construct temp novel gene db
talon.make_temp_novel_gene_table(cursor, "toy_build")
chrom = "chr1"
start = 1000
end = 1
edge_IDs = [run_info.edge + 1]
positions = [1000, 900, 100, 1]
vertex_IDs = [5, 2]
strand = "-"
anti_strand = "+"
v_novelty = (0, 0, 0, 0)
# Find antisense match
gene_ID, transcript_ID, gene_novelty, transcript_novelty, start_end_info = \
talon.process_spliced_antisense(chrom, positions,
strand, edge_IDs,
vertex_IDs,
transcript_dict,
gene_starts,
gene_ends,
edge_dict, locations,
vertex_2_gene, run_info,
cursor)
#anti_gene_ID = talon.find_gene_match_on_vertex_basis(vertex_IDs,
# anti_strand,
# vertex_2_gene)
correct_gene_ID = fetch_correct_ID("TG1", "gene", cursor)
anti_gene_ID = gene_novelty[-1][-1]
assert anti_gene_ID == correct_gene_ID
assert start_end_info["vertex_IDs"] == [6, 5, 2, 1]
conn.close()
def test_abundance(self):
""" Try to add abundance entries to database in batches
"""
conn, cursor = get_db_cursor()
build = "toy_build"
edge_dict = talon.make_edge_dict(cursor)
run_info = talon.init_run_info(cursor, build)
abundance = [(1, "test", 5), (2, "test", 1), (3, "test", 2)]
batch_size = 2
talon.batch_add_abundance(cursor, abundance, batch_size)
# Test if items are there
query = "SELECT * FROM abundance"
cursor.execute(query)
assert len(cursor.fetchall()) == 3
conn.close()
def test_all_known_edges(self):
""" Example where the toy transcript database contains matches for all
vertices.
"""
conn, cursor = get_db_cursor()
build = "toy_build"
edge_dict = talon.make_edge_dict(cursor)
run_info = talon.init_run_info(cursor, build)
conn.close()
chrom = "chr1"
vertex_IDs = [1, 2, 3, 4, 5, 6]
strand = "+"
edge_IDs, novelty = talon.match_all_transcript_edges(
vertex_IDs, strand, edge_dict, run_info)
assert edge_IDs == (1, 2, 3, 4, 5)
assert novelty == (0, 0, 0, 0, 0)
def test_transcript_annot(self):
""" Try to add transcript annotation entries to database in batches
"""
conn, cursor = get_db_cursor()
build = "toy_build"
edge_dict = talon.make_edge_dict(cursor)
run_info = talon.init_run_info(cursor, build)
annot = [(1, "toy", "TALON", "status", "NOVEL"),
(2, "toy", "TALON", "status", "NOVEL")]
batch_size = 2
talon.batch_add_annotations(cursor, annot, "transcript", batch_size)
# Test if items are there
query = "SELECT * FROM transcript_annotations WHERE value = 'NOVEL'"
cursor.execute(query)
assert len(cursor.fetchall()) == 2
conn.close()
def test_observed(self):
""" Try to add observed entries to database in batches
"""
conn, cursor = get_db_cursor()
build = "toy_build"
edge_dict = talon.make_edge_dict(cursor)
run_info = talon.init_run_info(cursor, build)
observed = [(1, 1, 1, "read1", "test", 1, 2, 1, 1, 0, 0, 100),
(2, 1, 1, "read2", "test", 1, 2, 1, 1, 0, 0, 100),
(3, 1, 1, "read3", "test", 1, 2, 1, 1, 0, 0, 100)]
batch_size = 1
talon.batch_add_observed(cursor, observed, batch_size)
# Test if items are there
query = "SELECT * FROM observed"
cursor.execute(query)
assert len(cursor.fetchall()) == 3
conn.close()
def test_gene_update(self):
""" Try to add novel gene entries to database while ignoring duplicates
"""
conn, cursor = get_db_cursor()
build = "toy_build"
edge_dict = talon.make_edge_dict(cursor)
run_info = talon.init_run_info(cursor, build)
talon.make_temp_novel_gene_table(cursor, build)
talon.create_gene("chr4", 1, 1000, "+", cursor, run_info)
talon.add_genes(cursor)
# Test if gene with ID 6 is there, but make sure we didn't add
# duplicates of the other genes
query = "SELECT * FROM genes"
gene_IDs = [x['gene_ID'] for x in cursor.execute(query)]
assert 7 in gene_IDs
assert len(gene_IDs) == 7
conn.close()
def test_overlap_but_no_vertex_match(self):
""" Example where the transcript is short, so it overlaps the
annotated transcript but is not an accepted match.
the start should get assigned to the annotated end, but the end is
novel """
conn, cursor = get_db_cursor()
build = "toy_build"
talon.make_temp_novel_gene_table(cursor, build)
talon.make_temp_monoexonic_transcript_table(cursor, build)
edge_dict = talon.make_edge_dict(cursor)
location_dict = talon.make_location_dict(build, cursor)
run_info = talon.init_run_info(cursor, build)
transcript_dict = talon.make_transcript_dict(cursor, build)
vertex_2_gene = talon.make_vertex_2_gene_dict(cursor)
gene_starts, gene_ends = talon.make_gene_start_and_end_dict(cursor, build)
tot_vertices = len(vertex_2_gene)
query = """ SELECT COUNT(*) FROM temp_monoexon """
tot_monoexonic = cursor.execute(query).fetchone()[0]
chrom = "chr4"
strand = "-"
positions = ( 2500, 2000 )
annotation = talon.identify_monoexon_transcript(chrom, positions,
strand, cursor,
location_dict, edge_dict,
transcript_dict, vertex_2_gene,
gene_starts, gene_ends, run_info)
correct_gene_ID = fetch_correct_ID("TG6", "gene", cursor)
print(annotation['start_vertex'])
print(annotation['end_vertex'])
assert annotation['gene_ID'] == correct_gene_ID
assert annotation['start_delta'] == None
assert annotation['end_delta'] == None
# Now check if the transcript got added to the right data structures
assert len(vertex_2_gene) == tot_vertices + 2
assert cursor.execute(query).fetchone()[0] == tot_monoexonic + 1
conn.close()
def test_edge_update(self):
""" Try to add novel exons and introns. """
conn, cursor = get_db_cursor()
build = "toy_build"
edge_dict = talon.make_edge_dict(cursor)
run_info = talon.init_run_info(cursor, build)
orig_n_edges = run_info.edge
talon.create_edge(2, 1, "exon", "-", edge_dict, run_info)
batch_size = 10
talon.batch_add_edges(cursor, edge_dict, batch_size)
# Test if the edge table has the correct number of edges now
query = "SELECT * FROM edge"
cursor.execute(query)
edge_IDs = [x['edge_ID'] for x in cursor.fetchall()]
assert orig_n_edges + 1 in edge_IDs
assert len(edge_IDs) == orig_n_edges + 1
conn.close()
def test_no_match(self):
""" Example with no ISM match """
conn, cursor = get_db_cursor()
build = "toy_build"
edge_dict = talon.make_edge_dict(cursor)
location_dict = talon.make_location_dict(build, cursor)
run_info = talon.init_run_info(cursor, build)
transcript_dict = talon.make_transcript_dict(cursor, build)
gene_starts, gene_ends = talon.make_gene_start_and_end_dict(
cursor, build)
chrom = "chr1"
strand = "+"
positions = [1, 100, 900, 1000]
edge_IDs = [200]
vertex_IDs = [2, 5]
v_novelty = [0, 0]
all_matches = talon.search_for_ISM(edge_IDs, transcript_dict)
assert all_matches == None
conn.close()
def test_genomic(self):
""" Example where the transcript overlaps a gene but contains no known
splice vertices
"""
conn, cursor = get_db_cursor()
build = "toy_build"
edge_dict = talon.make_edge_dict(cursor)
location_dict = talon.make_location_dict(build, cursor)
run_info = talon.init_run_info(cursor, build)
transcript_dict = talon.make_transcript_dict(cursor, build)
vertex_2_gene = talon.make_vertex_2_gene_dict(cursor)
gene_starts, gene_ends = talon.make_gene_start_and_end_dict(
cursor, build)
# Construct temp novel gene db
talon.make_temp_novel_gene_table(cursor, "toy_build")
chrom = "chr1"
positions = [1000, 950, 700, 600]
edge_IDs = [run_info.edge + 1, run_info.edge + 2]
vertex_IDs = [run_info.vertex + 1, run_info.vertex + 2]
strand = "-"
gene_ID, transcript_ID, gene_novelty, transcript_novelty, start_end_info = \
talon.process_remaining_mult_cases(chrom, positions,
strand, edge_IDs,
vertex_IDs,
transcript_dict,
gene_starts, gene_ends,
edge_dict, location_dict,
vertex_2_gene, run_info,
cursor)
correct_gene_ID = fetch_correct_ID("TG3", "gene", cursor)
assert gene_ID == correct_gene_ID
assert transcript_dict[frozenset(start_end_info["edge_IDs"])] != None
assert gene_novelty == []
assert transcript_novelty[-1][-2] == "genomic_transcript"
conn.close()
def test_intergenic(self):
""" Example where the transcript is an NIC match to an existing one by
virtue of a new splice donor.
"""
conn, cursor = get_db_cursor()
build = "toy_build"
edge_dict = talon.make_edge_dict(cursor)
location_dict = talon.make_location_dict(build, cursor)
run_info = talon.init_run_info(cursor, build)
transcript_dict = talon.make_transcript_dict(cursor, build)
vertex_2_gene = talon.make_vertex_2_gene_dict(cursor)
gene_starts, gene_ends = talon.make_gene_start_and_end_dict(
cursor, build)
correct_gene_ID = run_info.genes + 1
# Construct temp novel gene db
talon.make_temp_novel_gene_table(cursor, "toy_build")
chrom = "chrX"
positions = [1, 100, 900, 1000]
edge_IDs = [run_info.edge + 1, run_info.edge + 2]
vertex_IDs = [run_info.vertex + 1, run_info.vertex + 2]
strand = "+"
gene_ID, transcript_ID, gene_novelty, transcript_novelty, start_end_info = \
talon.process_remaining_mult_cases(chrom, positions,
strand, edge_IDs,
vertex_IDs,
transcript_dict,
gene_starts, gene_ends,
edge_dict, location_dict,
vertex_2_gene, run_info,
cursor)
assert gene_ID == correct_gene_ID
assert transcript_dict[frozenset(start_end_info["edge_IDs"])] != None
assert gene_novelty[0][-2] == "intergenic_novel"
conn.close()
def test_NIC_with_all_known_edges(self):
""" Test case derived from a real mouse Map2k4 read. All of edges are
known (except 3'), yet the read is NIC not FSM/ISM """
conn = sqlite3.connect("scratch/Map2k4.db")
conn.row_factory = sqlite3.Row
cursor = conn.cursor()
build = "mm10"
talon.make_temp_novel_gene_table(cursor, build)
edge_dict = talon.make_edge_dict(cursor)
location_dict = talon.make_location_dict(build, cursor)
run_info = talon.init_run_info(cursor, build)
transcript_dict = talon.make_transcript_dict(cursor, build)
vertex_2_gene = talon.make_vertex_2_gene_dict(cursor)
gene_starts, gene_ends = talon.make_gene_start_and_end_dict(
cursor, build)
chrom = "chr11"
strand = "-"
positions = [
65788254, 65788136, 65775765, 65775733, 65756371, 65756269,
65735366, 65735192, 65719603, 65719484, 65712297, 65712178,
65709983, 65709932, 65707111, 65706984, 65696365, 65696288,
65693570, 65693422, 65691773, 65691728, 65690804, 65689322
]
annotation = talon.identify_transcript(chrom, positions, strand,
cursor, location_dict,
edge_dict, transcript_dict,
vertex_2_gene, gene_starts,
gene_ends, run_info)
assert annotation['gene_ID'] == 1
assert annotation['transcript_ID'] == 8
novelty_types = [x[-2] for x in annotation['transcript_novelty']]
assert "NIC_transcript" in novelty_types
conn.close()
