def testEquals(self):
mite = Mite("ABC", "4", "+", "132", "154", "NULL", "Unclassified",
"Unclassified", "3")
gene = Gene("gABC", "4", "+", "132", "154", "5", "3")
pair1 = MiteGenePair(mite, gene)
pair2 = MiteGenePair(mite, gene)
self.assertEqual(pair1, pair2)
def testToRowMiteInGene(self):
mite = Mite("ABC", "4", "+", "132", "154", "NULL", "Unclassified",
"Unclassified", "3")
gene = Gene("gABC", "4", "+", "120", "160", "5", "3")
pair = MiteGenePair(mite, gene)
csv_row_string = pair.to_csv_row()
expected_string = "ABC,4,+,132,154,3,,,,gABC,4,+,120,160,3,5,0"
self.assertEqual(expected_string, csv_row_string)
def testNotEquals(self):
mite1 = Mite("ABC", "4", "+", "132", "154", "NULL", "Unclassified",
"Unclassified", "3")
gene1 = Gene("gABC", "4", "+", "132", "154", "5", "3")
mite2 = Mite("ABC", "3", "+", "132", "154", "NULL", "Unclassified",
"Unclassified", "3")
gene2 = Gene("gABC", "4", "+", "132", "154", "5", "3")
pair1 = MiteGenePair(mite1, gene1)
pair2 = MiteGenePair(mite2, gene2)
self.assertNotEqual(pair1, pair2)
def test_pair_mites_correctly_separates_by_genome_location(self):
mite1 = Mite("ABC1", "1", "+", "1", "10", "NULL", "AA", "A", "1")
mite2 = Mite("ABC2", "1", "-", "21", "30", "NULL", "AA", "A", "1")
mite3 = Mite("ABC3", "2", "+", "41", "50", "NULL", "AA", "A", "1")
mite4 = Mite("ABC4", "2", "-", "61", "70", "NULL", "AA", "A", "1")
mite5 = Mite("ABC5", "1", "+", "81", "90", "NULL", "AA", "A", "2")
mite6 = Mite("ABC6", "1", "-", "101", "110", "NULL", "AA", "A", "2")
mite7 = Mite("ABC7", "2", "+", "121", "130", "NULL", "AA", "A", "2")
mite8 = Mite("ABC8", "2", "-", "141", "150", "NULL", "AA", "A", "2")
gene1 = Gene("gABC1", "1", "+", "126", "135", "5", "1")
gene2 = Gene("gABC2", "1", "-", "146", "155", "5", "1")
gene3 = Gene("gABC3", "2", "+", "6", "15", "5", "1")
gene4 = Gene("gABC4", "2", "-", "26", "35", "5", "1")
gene5 = Gene("gABC5", "1", "+", "46", "55", "5", "2")
gene6 = Gene("gABC6", "1", "-", "66", "75", "5", "2")
gene7 = Gene("gABC7", "2", "+", "86", "95", "5", "2")
gene8 = Gene("gABC8", "2", "-", "106", "115", "5", "2")
mites = [mite1, mite2, mite3, mite4, mite5, mite6, mite7, mite8]
genes = [gene1, gene2, gene3, gene4, gene5, gene6, gene7, gene8]
pairs = self.pairer.pair_mites_with_genes(mites, genes)
expected_pairs = [
MiteGenePair(mite1, gene1),
MiteGenePair(mite2, gene1),
MiteGenePair(mite3, gene4),
MiteGenePair(mite4, gene4),
MiteGenePair(mite5, gene6),
MiteGenePair(mite6, gene6),
MiteGenePair(mite7, gene8),
MiteGenePair(mite8, gene8)
]
self.assertEqual(pairs, expected_pairs)
def main():
mites_path = "../../data/all_mites.csv"
genes_path = "../../data/all_genes.csv"
mites_parser = MiteCSVParser()
genes_parser = GeneCSVParser()
pairer = MitePairer()
start_time = time.time()
mites = mites_parser.parse_csv(mites_path)
end_time = time.time()
print("Mite CSV parsing took " + str(end_time - start_time) + "s")
start_time = time.time()
genes = genes_parser.parse_csv(genes_path)
end_time = time.time()
print("Gene CSV parsing took " + str(end_time - start_time) + "s")
start_time = time.time()
pairs = pairer.pair_mites_with_genes(mites, genes)
end_time = time.time()
print("Finding mite-gene pairs took " + str(end_time - start_time) + "s")
start_time = time.time()
output_path = "../../results/mite_gene_pairs.csv"
output_file = open(output_path, mode="w+")
output_file.write(MiteGenePair.csv_headers() + "\n")
for pair in pairs:
output_file.write(pair.to_csv_row() + "\n")
output_file.close()
end_time = time.time()
print("Writing mite-gene pairs took " + str(end_time - start_time) + "s")
def pair_mites_with_genes(self, mites, genes):
pairs = []
indexer = GeneticElementIndex()
all_elements = mites + genes
index = indexer.index_elements(all_elements, by_name=True)
indexed_list = indexer.index_elements(all_elements, by_name=False)
for genome_key in indexed_list:
genome = indexed_list[genome_key]
for chromosome_key in genome:
elements = genome[chromosome_key].copy()
elements.sort(key=lambda e: e.start)
for idx in range(len(elements)):
element = elements[idx]
if element.type != CONSTANTS.mite_type:
continue
correlated_gene_id = element.correlated_gene
index_bin = index[genome_key][chromosome_key]
if correlated_gene_id is not None and correlated_gene_id in index_bin and index_bin[
correlated_gene_id].type == CONSTANTS.gene_type:
correlated_gene = index_bin[correlated_gene_id]
pair = MiteGenePair(element, correlated_gene)
pairs.append(pair)
continue
pair = self._get_closest_mite_gene_pair(
elements, element, idx)
if pair is not None:
pairs.append(pair)
return pairs
def test_pair_mites_with_genes_results_independent_of_strand(self):
mite1 = Mite("ABC", "4", "+", "132", "154", "NULL", "AA", "A", "3")
mite2 = Mite("ABC1", "4", "+", "0", "15", "NULL", "AA", "A", "3")
mite3 = Mite("ABC2", "4", "-", "190", "200", "NULL", "AA", "A", "3")
mite4 = Mite("ABC3", "4", "-", "50", "65", "NULL", "AA", "A", "3")
gene1 = Gene("gABC", "4", "-", "100", "115", "5", "3")
gene2 = Gene("gBBC", "4", "+", "160", "170", "5", "3")
mites = [mite1, mite2, mite3, mite4]
genes = [gene1, gene2]
pairs = self.pairer.pair_mites_with_genes(mites, genes)
expected_pairs = [
MiteGenePair(mite2, gene1),
MiteGenePair(mite4, gene1),
MiteGenePair(mite1, gene2),
MiteGenePair(mite3, gene2)
]
self.assertEqual(pairs, expected_pairs)
def test_pair_mites_uses_heuristics_if_correlated_not_exist(self):
#go back to heuristics if we can't find the correlated gene
mite1 = Mite("ABC", "4", "+", "132", "154", "ABC", "AA", "A", "3")
gene1 = Gene("gDBC", "4", "+", "0", "10", "5", "3")
gene2 = Gene("gBBC", "4", "+", "155", "170", "5", "3")
gene3 = Gene("gCBC", "4", "+", "100", "143", "5", "3")
mites = [mite1]
genes = [gene1, gene2, gene3]
pairs = self.pairer.pair_mites_with_genes(mites, genes)
expected_pairs = [MiteGenePair(mite1, gene3)]
self.assertEqual(pairs, expected_pairs)
def test_pair_mites_uses_correlated_gene_if_provided(self):
#we assume that correlated_gene was filled for a reason so we ignore our distance heuristic or any other heuristic
mite1 = Mite("ABC", "4", "+", "132", "154", "gABC", "AA", "A", "3")
gene1 = Gene("gABC", "4", "+", "0", "10", "5", "3")
gene2 = Gene("gBBC", "4", "+", "155", "170", "5", "3")
gene3 = Gene("gCBC", "4", "+", "100", "143", "5", "3")
mites = [mite1]
genes = [gene1, gene2, gene3]
pairs = self.pairer.pair_mites_with_genes(mites, genes)
expected_pairs = [MiteGenePair(mite1, gene1)]
self.assertEqual(pairs, expected_pairs)
def test_pair_mites_prefers_nested_elements_input_order_doesnt_matter(
self):
#sanity check
mite1 = Mite("ABC", "4", "+", "132", "154", "NULL", "AA", "A", "3")
gene1 = Gene("gABC", "4", "+", "100", "143", "5", "3")
gene2 = Gene("gBBC", "4", "+", "155", "170", "5", "3")
mites = [mite1]
genes = [gene2, gene1]
pairs = self.pairer.pair_mites_with_genes(mites, genes)
expected_pairs = [MiteGenePair(mite1, gene1)]
self.assertEqual(pairs, expected_pairs)
def test_pair_mites_prefers_nested_elements(self):
#we assume that if a mite is nested in a gene (partially or completely), then it's associated with the gene
#so we pair the two elements together rather than another gene that's not nested but may be closer
mite1 = Mite("ABC", "4", "+", "132", "154", "NULL", "AA", "A", "3")
gene1 = Gene("gABC", "4", "+", "100", "143", "5", "3")
gene2 = Gene("gBBC", "4", "+", "155", "170", "5", "3")
mites = [mite1]
genes = [gene1, gene2]
pairs = self.pairer.pair_mites_with_genes(mites, genes)
expected_pairs = [MiteGenePair(mite1, gene1)]
self.assertEqual(pairs, expected_pairs)
def _get_closest_mite_gene_pair(self, elements, mite, idx):
left_gene = self._nearest_gene_left_of_idx(elements, idx)
right_gene = self._nearest_gene_right_of_idx(elements, idx)
if left_gene is not None and right_gene is None:
pair = MiteGenePair(mite, left_gene)
elif left_gene is None and right_gene is not None:
pair = MiteGenePair(mite, right_gene)
elif left_gene is not None and right_gene is not None:
if self._is_contained_in(left_gene, mite):
pair = MiteGenePair(mite, left_gene)
elif self._is_contained_in(right_gene, mite):
pair = MiteGenePair(mite, right_gene)
else:
left_distance = mite.start - left_gene.end
right_distance = right_gene.start - mite.end
if left_distance < right_distance:
pair = MiteGenePair(mite, left_gene)
else:
pair = MiteGenePair(mite, right_gene)
else:
pair = None
return pair
def testHeaders(self):
headers = MiteGenePair.csv_headers()
expected_string = "mite_name,mite_chromosome,mite_strand,mite_start,mite_end,mite_genome_id," \
"mite_correlated_gene,mite_family,mite_superfamily,gene_name,gene_chromosome,gene_strand," \
"gene_start,gene_end,gene_genome_id,gene_exon_count,distance"
self.assertEqual(expected_string, headers)