def filter_gff(hits): hits_to_keep = [] hits = sorted(hits,key= lambda x: int(x[0][3])) #Get only the features annotated as genes gene_annotations = [x for y in hits for x in y if x[2] == 'gene'] #Get the start,end, and score for each gene annotation range_list = [(int(x[3]),int(x[4])) for x in gene_annotations] kept_indicies = range_connectivity(range_list) if len(kept_indicies) > 1: overlapping_indicies = [] non_overlapping_indicies = [] for ix in range(len(kept_indicies)-1): #print range_list[ix],range_list[ix+1],tuple_overlap(range_list[ix],range_list[ix+1]) if tuple_overlap(range_list[ix],range_list[ix+1]): overlapping_indicies.append((kept_indicies[ix],kept_indicies[ix+1])) else: non_overlapping_indicies.append(kept_indicies[ix]) for pair in overlapping_indicies: if int(gene_annotations[pair[0]][5]) > int(gene_annotations[pair[1]][5]): non_overlapping_indicies.append(pair[0]) else: non_overlapping_indicies.append(pair[1]) if not tuple_overlap(range_list[-2],range_list[-1]): non_overlapping_indicies.append(kept_indicies[-1]) return [hits[x] for x in non_overlapping_indicies] else: return [hits[x] for x in kept_indicies]
def filter_gff(hits,merge=True):
hits_to_keep = []
hits = sorted(hits,key= lambda x: int(x[0][3]))
#Get only the features annotated as genes
gene_annotations = [x for y in hits for x in y if x[2] == 'gene']
#Get the start,end, and score for each gene annotation
range_list = [(int(x[3]),int(x[4])) for x in gene_annotations]
#print(range_list)
kept_indicies = range_connectivity(range_list)
kept_range_list = [range_list[x] for x in kept_indicies]
#print(kept_indicies)
if len(kept_indicies) > 1:
overlapping_indicies = []
non_overlapping_indicies = []
for ix in range(len(kept_indicies)-1):
#print range_list[ix],range_list[ix+1],tuple_overlap(range_list[ix],range_list[ix+1])
if tuple_overlap(kept_range_list[ix],kept_range_list[ix+1]):
if kept_indicies[ix] not in overlapping_indicies:
# if not tuple_overlap(kept_range_list[ix-1],kept_range_list[ix]):
overlapping_indicies.append(kept_indicies[ix])
if kept_indicies[ix+1] not in overlapping_indicies:
overlapping_indicies.append(kept_indicies[ix+1])
else:
non_overlapping_indicies.append(kept_indicies[ix])
#print overlapping_indicies
if overlapping_indicies:
best_score = score_filter([hits[x] for x in overlapping_indicies])
if best_score:
non_overlapping_indicies.append(best_score)
else:
if merge:
#merge the gene first
gene_anno = []
cds_anno = []
exon_anno = []
intron_anno = []
similarity_anno = []
misc_anno = []
#print(hits)
for x in hits:
for l in x:
if l[2] == 'gene':
gene_anno.append(l)
elif l[2] == "cds":
cds_anno.append(l)
elif l[2] == "exon":
exon_anno.append(l)
elif l[2] == "intron":
intron_anno.append(l)
elif l[2] == "similarity":
similarity_anno.append(l)
else:
misc_anno.append(l)
print("Merging {} annotations".format(len(gene_anno)))
joined_gene = join_zones(gene_anno)
joined_cds = join_zones(cds_anno)
joined_exon = join_zones(exon_anno)
joined_similarity = join_zones(similarity_anno)
joined_hit = [joined_gene,joined_cds,joined_exon]
if intron_anno:
joined_intron = join_zones(intron_anno)
joined_hit.append(joined_intron)
joined_hit.append(joined_similarity)
joined_hit += misc_anno
hits.append(joined_hit)
kept_indicies.append(len(hits)-1)
non_overlapping_indicies.append(len(kept_indicies)-1)
else:
longest = longest_hit([hits[x] for x in overlapping_indicies])
if longest:
non_overlapping_indicies.append(longest)
# for pair in overlapping_indicies:
# if int(gene_annotations[pair[0]][5]) > int(gene_annotations[pair[1]][5]):
# non_overlapping_indicies.append(pair[0])
# else:
# non_overlapping_indicies.append(pair[1])
# if not tuple_overlap(range_list[-2],range_list[-1]):
# non_overlapping_indicies.append(kept_indicies[-1])
#print kept_indicies
#print non_overlapping_indicies
return [hits[kept_indicies[x]] for x in sorted(non_overlapping_indicies)]#.sort()]
else:
return [hits[x] for x in kept_indicies]
def filter_gff(hits, merge=True):
hits_to_keep = []
hits = sorted(hits, key=lambda x: int(x[0][3]))
#Get only the features annotated as genes
gene_annotations = [x for y in hits for x in y if x[2] == 'gene']
#Get the start,end, and score for each gene annotation
range_list = [(int(x[3]), int(x[4])) for x in gene_annotations]
#print(range_list)
kept_indicies = range_connectivity(range_list)
kept_range_list = [range_list[x] for x in kept_indicies]
#print(kept_indicies)
if len(kept_indicies) > 1:
overlapping_indicies = []
non_overlapping_indicies = []
for ix in range(len(kept_indicies) - 1):
#print range_list[ix],range_list[ix+1],tuple_overlap(range_list[ix],range_list[ix+1])
if tuple_overlap(kept_range_list[ix], kept_range_list[ix + 1]):
if kept_indicies[ix] not in overlapping_indicies:
# if not tuple_overlap(kept_range_list[ix-1],kept_range_list[ix]):
overlapping_indicies.append(kept_indicies[ix])
if kept_indicies[ix + 1] not in overlapping_indicies:
overlapping_indicies.append(kept_indicies[ix + 1])
else:
non_overlapping_indicies.append(kept_indicies[ix])
#print overlapping_indicies
if overlapping_indicies:
best_score = score_filter([hits[x] for x in overlapping_indicies])
if best_score:
non_overlapping_indicies.append(best_score)
else:
if merge:
#merge the gene first
gene_anno = []
cds_anno = []
exon_anno = []
intron_anno = []
similarity_anno = []
misc_anno = []
#print(hits)
for x in hits:
for l in x:
if l[2] == 'gene':
gene_anno.append(l)
elif l[2] == "cds":
cds_anno.append(l)
elif l[2] == "exon":
exon_anno.append(l)
elif l[2] == "intron":
intron_anno.append(l)
elif l[2] == "similarity":
similarity_anno.append(l)
else:
misc_anno.append(l)
print("Merging {} annotations".format(len(gene_anno)))
joined_gene = join_zones(gene_anno)
joined_cds = join_zones(cds_anno)
joined_exon = join_zones(exon_anno)
joined_similarity = join_zones(similarity_anno)
joined_hit = [joined_gene, joined_cds, joined_exon]
if intron_anno:
joined_intron = join_zones(intron_anno)
joined_hit.append(joined_intron)
joined_hit.append(joined_similarity)
joined_hit += misc_anno
hits.append(joined_hit)
kept_indicies.append(len(hits) - 1)
non_overlapping_indicies.append(len(kept_indicies) - 1)
else:
longest = longest_hit(
[hits[x] for x in overlapping_indicies])
if longest:
non_overlapping_indicies.append(longest)
# for pair in overlapping_indicies:
# if int(gene_annotations[pair[0]][5]) > int(gene_annotations[pair[1]][5]):
# non_overlapping_indicies.append(pair[0])
# else:
# non_overlapping_indicies.append(pair[1])
# if not tuple_overlap(range_list[-2],range_list[-1]):
# non_overlapping_indicies.append(kept_indicies[-1])
#print kept_indicies
#print non_overlapping_indicies
try:
return_value = []
for x in sorted(non_overlapping_indicies):
try:
return_value.append(hits[kept_indicies[x]])
except Exception, e:
print(e)
print("x is: {}".format(x))
print("kept_indicies is: {}".format(kept_indicies))
print("hits is: {}".format(hits))
return return_value
except Exception, e:
print(e)