def remove_duplicates(self,args):
"""
- First create a table of those that overlap
- Then create merged entries based on the overlap matrix
"""
if(not self.genes_spanning_left_junction or not self.genes_spanning_right_junction):
raise Exception("Gene annotations on dataset '"+self.name+"' were not found")
else:
old_count = len(self)
if(self.name.find("vs.") == -1):
self.logger.info("Duplication removal: "+self.name+" ("+str(old_count)+" fusions)")
unique_fusions = []
if(args.matching_method in ["overlap","subset","egm"]):
from CompareFusionsBySpanningGenes import CompareFusionsBySpanningGenes
overlap = CompareFusionsBySpanningGenes(False,False,args)
else:
raise Exception("Unknown overlap method for removing duplicates: '"+args.matching_method+"' for dataset "+self.name)
stats_duplicates = 0
stats_non_gene_spanning = 0
fusions_to_add = []
for chromosome_left in self.index.items():
for chromosome_right in chromosome_left[1].items():
all_fusions = chromosome_right[1]
n = len(all_fusions)
queue = range(n)
while(len(queue) > 0):
duplicates = []
for i in queue:
fusion_1 = all_fusions[i]
if(fusion_1):
is_duplicate = False
if(len(fusion_1.get_annotated_genes_left()) == 0 or len(fusion_1.get_annotated_genes_right()) == 0):
stats_non_gene_spanning += 1
all_fusions[i] = False
else:
for j in range(i+1,n):
fusion_2 = all_fusions[j]
if(fusion_2):
match = overlap.match_fusions(fusion_1,fusion_2,False)
if(match):
fusion_1 = match
all_fusions[i] = match
all_fusions[j] = False
is_duplicate = True
if(is_duplicate):
duplicates.append(i)
else:
unique_fusions.append(fusion_1)
queue = duplicates
for fusion in all_fusions:
if(fusion):
fusions_to_add.append(fusion)
self.flush()
for fusion in fusions_to_add:
self.add_fusion(fusion)
if(self.name.find("vs.") == -1):
self.logger.info("* Full: "+str(old_count))
self.logger.info("* Gene-spanning: "+str(old_count-stats_non_gene_spanning))
self.logger.info("* Unique: "+str(len(self)))
return len(self)
def remove_duplicates(self, args):
"""
- First create a table of those that overlap
- Then create merged entries based on the overlap matrix
"""
if (not self.genes_spanning_left_junction
or not self.genes_spanning_right_junction):
raise Exception("Gene annotations on dataset '" + self.name +
"' were not found")
else:
old_count = len(self)
if (self.name.find("vs.") == -1):
self.logger.info("Duplication removal: " + self.name + " (" +
str(old_count) + " fusions)")
unique_fusions = []
if (args.matching_method in ["overlap", "subset", "egm"]):
from CompareFusionsBySpanningGenes import CompareFusionsBySpanningGenes
overlap = CompareFusionsBySpanningGenes(False, False, args)
else:
raise Exception(
"Unknown overlap method for removing duplicates: '" +
args.matching_method + "' for dataset " + self.name)
stats_duplicates = 0
stats_non_gene_spanning = 0
fusions_to_add = []
for chromosome_left in self.index.items():
for chromosome_right in chromosome_left[1].items():
all_fusions = chromosome_right[1]
n = len(all_fusions)
queue = range(n)
while (len(queue) > 0):
duplicates = []
for i in queue:
fusion_1 = all_fusions[i]
if (fusion_1):
is_duplicate = False
if (len(fusion_1.get_annotated_genes_left(False))
== 0 or len(
fusion_1.get_annotated_genes_right(
False)) == 0):
stats_non_gene_spanning += 1
all_fusions[i] = False
else:
for j in range(i + 1, n):
fusion_2 = all_fusions[j]
if (fusion_2):
match = overlap.match_fusions(
fusion_1, fusion_2, False)
if (match):
merged_matches = fusion_1.matches | fusion_2.matches
fusion_1.matches = merged_matches
fusion_1.acceptor_donor_direction = match.acceptor_donor_direction
fusion_1.left_strand = match.left_strand
fusion_1.right_strand = match.right_strand
fusion_1.annotated_genes_left = match.annotated_genes_left
fusion_1.annotated_genes_right = match.annotated_genes_right
all_fusions[i] = fusion_1
all_fusions[j] = False
is_duplicate = True
match.prepare_deletion()
del (match)
if (is_duplicate):
duplicates.append(i)
else:
unique_fusions.append(fusion_1)
queue = duplicates
for fusion in all_fusions:
if (fusion):
fusions_to_add.append(fusion)
self.flush()
for fusion in fusions_to_add:
self.add_fusion(fusion)
if (self.name.find("vs.") == -1):
self.logger.debug("* Full: " + str(old_count))
self.logger.debug("* Gene-spanning: " +
str(old_count - stats_non_gene_spanning))
self.logger.debug("* Unique: " + str(len(self)))
return len(self)
