def variants_in_UTR(collection, feature_type="5'_UTR"):
UTR_record_list = []
for variant in collection:
if feature_type in variant.info_dict["Ftype"]:
UTR_record_list.append(variant)
return CollectionVCF(from_file=False,
record_list=UTR_record_list,
metadata=collection.metadata,
header=collection.header)
def extract_vcf(self):
vcf = CollectionVCF(metadata=self.metadata.vcf_metadata,
record_list=[],
header=self.metadata.vcf_header,
samples=self.metadata.samples,
from_file=False)
for record in self:
"""
print(record)
print(type(record))
print(record.records)
print(type(record.records))
"""
vcf = vcf + record.records
return vcf
def variants_start_end(collection,
left,
record_dict,
min_five_utr_len=10,
skip_nonintergenic_variants=False):
pre_UTR_record_list = []
for record_id in record_dict:
for feature in record_dict[record_id].features:
if feature.type != "gene":
continue
for sub_feature in feature.sub_features:
if sub_feature.type == "five_prime_UTR" and len(
sub_feature) >= min_five_utr_len:
break
else:
continue
for sub_feature in feature.sub_features:
strand = sub_feature.strand
if sub_feature.type == "five_prime_UTR":
five_UTR_start = sub_feature.location.start + 1 if strand == +1 else sub_feature.location.end
pre_UTR_start = five_UTR_start - left if strand == +1 else five_UTR_start + 1
pre_UTR_end = five_UTR_start - 1 if strand == +1 else five_UTR_start + left
for variant in collection:
if record_id != variant.chrom:
continue
if skip_nonintergenic_variants and variant.info_dict[
"Ftype"] != ["igc"]:
continue
if pre_UTR_start <= variant.pos <= pre_UTR_end:
pre_UTR_record_list.append(variant)
pre_UTR_record_list[-1].info_dict["Fstrand"] = [
"P"
] if strand == +1 else ["M"]
relative_position = (variant.pos -
five_UTR_start) * strand
if relative_position > 0:
print(pre_UTR_start, pre_UTR_end,
five_UTR_start)
print(variant)
print(sub_feature)
return CollectionVCF(from_file=False,
record_list=pre_UTR_record_list,
metadata=collection.metadata,
header=collection.header)
help="End histogramm file prefix",
default="end_histogramm")
args = parser.parse_args()
if args.both is not None:
args.left = args.both
args.right = args.both
if (not args.right) and (not args.left):
raise ValueError("Both left and right regions were not set")
with open(args.gff, "r") as in_fd:
record_dict = dict([(record.id, record) for record in GFF.parse(in_fd)])
variants = CollectionVCF(from_file=True, vcf_file=args.vcf)
gene_variants_positions = []
all_variant_start_positions = []
all_variant_end_positions = []
print(args.left)
print(args.right)
for record_id in record_dict:
for feature in record_dict[record_id].features:
if feature.type != "gene":
continue
#print(feature.sub_features)
for sub_feature in feature.sub_features:
if sub_feature.type != "CDS":
continue
chrom = record_id
strand = sub_feature.strand
annotations_dict = {}
annotation_synonym_dict = {
"three_prime_UTR": "3'_UTR",
"five_prime_UTR": "5'_UTR",
"snoRNA": "ncRNA",
"snRNA": "ncRNA"
}
annotation_black_list = [
"region", "ARS", "long_terminal_repeat", "noncoding_exon", "intron",
"repeat_region"
]
with open(gff_file) as gff_fd:
for record in GFF.parse(gff_fd):
annotations_dict[record.id] = record
bad_region_dict = {}
with open(bad_regions_file) as gff_fd:
for record in GFF.parse(gff_fd):
bad_region_dict[record.id] = record
for sample_set_name in sample_set_names_list:
print("Handling %s" % sample_set_name)
os.chdir(workdir)
mutations = CollectionVCF(vcf_file=sample_set_name + ".vcf",
from_file=True)
mutations.find_location(annotations_dict,
use_synonym=True,
synonym_dict=annotation_synonym_dict,
feature_type_black_list=annotation_black_list)
mutations.write("%s_annotated.vcf" % sample_set_name)
import matplotlib.pyplot as plt
from matplotlib import rcParams
if __name__ == "__main__":
workdir = "/media/mahajrod/d9e6e5ee-1bf7-4dba-934e-3f898d9611c8/Data/LAN2xx/combined_vcf/clusters/all/pre_UTR_strandness/"
letter_list_part1 = ["A", "B", "C", "D", "E", "F", "G", "H", "I", "J"]
os.chdir(workdir)
sample_set_names_list = [
"PmCDA1_3d", "PmCDA1_sub1_3d", "PmCDA1_6d", "PmCDA1_sub1_6d"
]
rcParams.update({'font.size': 7})
plt.figure(1, dpi=300, figsize=(6, 6))
index = 1
for sample, letter in zip(sample_set_names_list, letter_list_part1):
collection = CollectionVCF(
from_file=True,
vcf_file=sample + "_pre_UTR_variants_only_intergenic_l_300.vcf")
sample_data = collection.count_strandness(
sample + "_pre_UTR_variants_only_intergenic_l_300_strandness")
plt.subplot(2, 2, index)
n_groups = 4
points = np.arange(n_groups)
bar_width = 0.35
C_values = sample_data["all"][0]
G_values = sample_data["all"][1]
rects1 = plt.bar(points, C_values, bar_width, color='b', label='C->T')
rects2 = plt.bar(points + bar_width,
G_values,
from matplotlib import rcParams
if __name__ == "__main__":
workdir = "/media/mahajrod/d9e6e5ee-1bf7-4dba-934e-3f898d9611c8/Data/LAN2xx/combined_vcf/clusters/all/all/"
letter_list_part1 = ["A", "B", "C", "D", "E", "F", "G", "H", "I", "J"]
os.chdir(workdir)
sample_set_names_list = ["PmCDA1_3d",
"PmCDA1_sub1_3d",
"PmCDA1_6d",
"PmCDA1_sub1_6d"
]
rcParams.update({'font.size': 7})
plt.figure(1, dpi=300, figsize=(6, 6))
index = 1
for sample, letter in zip(sample_set_names_list, letter_list_part1):
collection = CollectionVCF(from_file=True, vcf_file=sample + "_good.vcf")
sample_data = collection.count_strandness(sample + "_good_strandness")
plt.subplot(2, 2, index)
n_groups = 4
points = np.arange(n_groups)
bar_width = 0.35
C_values = sample_data["all"][0]
G_values = sample_data["all"][1]
rects1 = plt.bar(points, C_values, bar_width,
color='b',
label='C->T')
rects2 = plt.bar(points + bar_width, G_values, bar_width,
color='g',
os.chdir(workdir)
skip_genes_without_five_utr = False
left = 300
right = 300
bin_width = 5
bins = np.linspace(-left, right, ((left + right) / bin_width) + 1)
normed = True
max_start = 0
max_end = 0
for sample_set in sample_set_names_list:
vcf_file = "%s_good.vcf" % sample_set
#start_hist_prefix = "%s_start_hist_r_%i_l_%i" % (sample_set, right, left)
#end_hist_prefix = "%s_end_hist_r_%i_l_%i" % (sample_set, right, left)
#gene_variants = "%s_gene_variants_r_%i_l_%i.t" % (sample_set, right, left)
variants = CollectionVCF(from_file=True, vcf_file=vcf_file)
start_dict[sample_set], end_dict[sample_set], position_dict[sample_set] = \
variants.variants_start_end(left, right, record_dict, skip_genes_without_five_utr=skip_genes_without_five_utr)
length_dict[sample_set] = len(variants)
#print(start_dict[sample_set])
start_hist_dict[sample_set] = list(
np.histogram(start_dict[sample_set], bins=bins))
end_hist_dict[sample_set] = list(
np.histogram(end_dict[sample_set], bins=bins))
print(start_hist_dict[sample_set][0])
if normed:
start_hist_dict[sample_set][0] = start_hist_dict[sample_set][
0].astype(np.float32, copy=False)
end_hist_dict[sample_set][0] = end_hist_dict[sample_set][0].astype(
np.float32, copy=False)
start_hist_dict[sample_set][
bad_region_dict = {}
with open(bad_regions_file) as gff_fd:
for record in GFF.parse(gff_fd):
bad_region_dict[record.id] = record
for sample_set_name in sample_set_names_list:
print("Handling %s" % sample_set_name)
os.chdir(workdir)
os.system("mkdir -p %s" % sample_set_name)
os.chdir(sample_set_name)
os.system("mkdir -p %s %s" % (clustering_dir, rainfall_dir))
#os.system("pwd")
mutations = CollectionVCF(
vcf_file="../SNP_annotated_raw_vcf/%s_SNP.vcf" % sample_set_name,
from_file=True)
"""
mutations.rainfall_plot("%s_mutations" % (sample_set_name), ref_genome=reference, draw_gaps=True,
masked_regions=bad_region_dict)
"""
mutations.get_location(annotations_dict,
use_synonym=True,
synonym_dict=annotation_synonym_dict)
mutations.check_location(bad_regions)
mutations.check_by_ref_and_alt(ref_alt_variants["desaminases"], "DA")
annotation_black_list = [
"gene", "region", "ARS", "long_terminal_repeat", "noncoding_exon",
"intron", "repeat_region", "telomere", "gene_cassette",
gff_file = "/home/mahajrod/genetics/desaminases/data/LAN210_v0.10m/annotations/merged_annotations_Nagalakshmi_tranf_to_LAN210_v0.10m.gff3"
annotations_dict = {}
with open(gff_file) as gff_fd:
for record in GFF.parse(gff_fd):
annotations_dict[record.id] = record
for sample in samples_list:
print("Handling %s" % sample)
os.chdir(workdir)
os.chdir(sample)
if alignment_dir not in os.listdir("."):
continue
os.chdir(alignment_dir)
os.system("mkdir -p %s" % clustering_dir)
mutations = CollectionVCF(vcf_file=sample + suffix, from_file=True)
mutations.get_location(annotations_dict)
mutations.check_by_ref_and_alt(ref_alt_variants["desaminases"], "DA")
"""
for record in mutations:
print (record)
print(record.flags)
"""
#for record in mutations:
# print(record.description)
mutations.location_pie(annotation_black_list=[
"gene", "region", "ARS", "long_terminal_repeat"
],
figsize=(40, 40),
pie_filename="variant_location_pie.svg",
counts_filename="variant_location_counts.t")
def adjust(self,
border_limit=None,
min_size_to_adjust=2,
remove_border_subclusters=False,
remove_size_limit=1):
# adjusts cluster borders, returns list of new cluster records
# skip adjustment for clusters with 3 or less mutations
if (self.size < min_size_to_adjust) or (self.subclusters is None):
#return -1
return [self]
limit = border_limit if border_limit else len(self.subclusters)
for i in range(0, limit):
if self.subclusters[i] == self.subclusters[0]:
left_subcluster_end = i
else:
break
# exit if cluster doesnt have subclusters
if left_subcluster_end == len(self.subclusters) - 1:
#return 1
return [self]
for i in range(-1, -limit - 1, -1):
if self.subclusters[i] == self.subclusters[-1]:
right_subcluster_start = i
else:
break
if remove_border_subclusters:
start = left_subcluster_end + 1 if left_subcluster_end < remove_size_limit else 0
end = right_subcluster_start if right_subcluster_start >= -remove_size_limit else len(
self.subclusters)
new_left_cluster, new_right_cluster = None, None
if start > 0:
new_left_cluster = RecordCCF(
collection_vcf=CollectionVCF(
record_list=self.records.records[:start],
from_file=False),
subclusters=self.subclusters[:start],
from_records=True)
if end < len(self.subclusters):
new_right_cluster = RecordCCF(
collection_vcf=CollectionVCF(
record_list=self.records.records[end:],
from_file=False),
subclusters=self.subclusters[end:],
from_records=True)
"""
self.__init__(collection_vcf=CollectionVCF(record_list=self.records.records[start:end], from_file=False),
subclusters=self.subclusters[start:end], from_records=True)
"""
new_middle_cluster = RecordCCF(
collection_vcf=CollectionVCF(
record_list=self.records.records[start:end],
from_file=False),
subclusters=self.subclusters[start:end],
from_records=True)
"""
if new_left_cluster or new_right_cluster:
print("original")
print(self)
print("adjusted")
print(new_left_cluster)
print(new_middle_cluster)
print(new_right_cluster)
"""
cluster_list = [new_left_cluster] if new_left_cluster else []
cluster_list += [new_middle_cluster]
cluster_list += [new_right_cluster] if new_right_cluster else []
return cluster_list
pre_UTR_bins = left / bin_width
CDS_bins = np.linspace(0, right, right / bin_width + 1)
UTR_bins = 10
normed = True
max_start = 0
max_end = 0
skip_nonintergenic_variants = True
for sample_set in sample_set_names_list:
print("Handling %s" % sample_set)
vcf_file = "%s_good.vcf" % sample_set
#start_hist_prefix = "%s_start_hist_r_%i_l_%i" % (sample_set, right, left)
#end_hist_prefix = "%s_end_hist_r_%i_l_%i" % (sample_set, right, left)
#gene_variants = "%s_gene_variants_r_%i_l_%i.t" % (sample_set, right, left)
#variants, minus_variants = CollectionVCF(from_file=True, vcf_file=vcf_file).filter_by_expression("record.ref == 'C'") # C -> T variants
#variants, minus_variants = CollectionVCF(from_file=True, vcf_file=vcf_file).filter_by_expression("(record.ref == 'C' and record.info_dict['Fstrand'][0] == 'P') or (record.ref == 'G' and record.info_dict['Fstrand'][0] == 'M')") # nontranscribed thread
variants = CollectionVCF(from_file=True, vcf_file=vcf_file)
pre_UTR_positions[sample_set], UTR_positions[sample_set], CDS_positions[sample_set] = \
variants_start_end(variants, left, right, record_dict, min_five_utr_len=10,
skip_nonintergenic_variants=skip_nonintergenic_variants)
length_dict[sample_set] = len(variants)
#print(start_dict[sample_set])
pre_UTR_hist_dict[sample_set] = list(
np.histogram(pre_UTR_positions[sample_set], bins=pre_UTR_bins))
UTR_hist_dict[sample_set] = list(
np.histogram(UTR_positions[sample_set], bins=UTR_bins))
CDS_hist_dict[sample_set] = list(
np.histogram(CDS_positions[sample_set], bins=CDS_bins))
print("UTR")
print(UTR_positions[sample_set])
print("blablabla")
print(pre_UTR_hist_dict[sample_set][0])
__author__ = 'mahajrod'
import os
from Parsers.VCF import CollectionVCF
if __name__ == "__main__":
workdir = "/media/mahajrod/d9e6e5ee-1bf7-4dba-934e-3f898d9611c8/Data/LAN2xx/all"
os.chdir(workdir)
samples_list = sorted(os.listdir("."))
suffix = "_GATK_best_merged.vcf"
for sample in samples_list:
print("Handling %s" % sample)
os.chdir(workdir)
os.chdir(sample)
if "alignment_LAN210_v0.9m" not in os.listdir("."):
continue
os.chdir("alignment_LAN210_v0.9m")
mutations = CollectionVCF(vcf_file=sample + suffix, from_file=True)
print("Totaly %s mutations" % len(mutations))
mutations.test_thresholds(save_clustering=True,
testing_dir="testing_theshold_inconsistent")
combined_vcf_suffix = "_adjusted_cluster_mutations.vcf"
combined_3_vcf_suffix = "_adjusted_3+_cluster_mutations.vcf"
samples_dir = "/media/mahajrod/d9e6e5ee-1bf7-4dba-934e-3f898d9611c8/Data/LAN2xx/all/"
samples_subdir = "alignment_LAN210_v0.10m/"
sample_suffix = "_GATK_best_SNP.vcf"
workdir = "/media/mahajrod/d9e6e5ee-1bf7-4dba-934e-3f898d9611c8/Data/LAN2xx/combined_vcf/"
combined_subdir = "clustering/"
for sample_set in sample_set_names_list:
os.chdir(workdir)
os.chdir(sample_set)
os.system("mkdir -p per_sample_vcf")
sample_set_mutations = CollectionVCF(from_file=True,
vcf_file=combined_subdir +
sample_set + combined_vcf_suffix)
#sample_set_3_mutations = CollectionVCF(from_file=True, vcf_file=combined_subdir + sample_set + combined_3_vcf_suffix)
per_sample_mutations = {}
samples_list = sample_set_mutations.samples
print("Handling %s" % sample_set)
for sample in sample_set_mutations.samples:
print("Handling %s" % sample)
sample_mutations = CollectionVCF(from_file=True,
vcf_file=samples_dir + sample + "/" +
samples_subdir + sample +
sample_suffix)
per_sample_mutations[sample] = CollectionVCF(
metadata=sample_mutations.metadata,
record_list=None,
header=sample_mutations.header,
"three_prime_UTR": "3'_UTR",
"five_prime_UTR": "5'_UTR",
"snoRNA": "ncRNA",
"snRNA": "ncRNA"
}
with open(gff_file) as gff_fd:
for record in GFF.parse(gff_fd):
annotations_dict[record.id] = record
bad_region_dict = {}
with open(bad_regions_file) as gff_fd:
for record in GFF.parse(gff_fd):
bad_region_dict[record.id] = record
region = "chrX"
for sample_set in sample_set_names_list:
collection = CollectionVCF(from_file=True,
vcf_file="%s_good.vcf" % sample_set)
plt.figure(1, figsize=(15, 8))
ax0 = plt.subplot(212)
ax0.set_yscale('log', basey=2)
ax0.set_xlim(1, 745688)
#ax0.set_ylim(ymin=0)
region_reference_dict, maximum = rainfall_plot(
collection,
region,
base_colors=[],
facecolor="#D6D6D6",
ref_genome=reference,
draw_gaps=True,
masked_regions=bad_region_dict)
for reference in region_reference_dict[region]:
continue
os.chdir("alignment_LAN210_v0.9m")
mutations = CollectionVCF(vcf_file=sample + suffix,
from_file=True)
print("Totaly %s mutations" % len(mutations))
mutations.test_thresholds(extracting_method='distance', threshold=(50, 5000, 100),
testing_dir="testing_threshold")
"""
sample_set_names_list = [
"PmCDA1_3d", "HAP", "PmCDA1_sub1_3d", "PmCDA1_6d", "HAP_sub1",
"PmCDA1_sub1_6d", "A1_3d", "A1_6d", "A3G_3d", "AID_3d", "AID_6d"
]
suffix = "_SNP.vcf"
workdir = "/media/mahajrod/d9e6e5ee-1bf7-4dba-934e-3f898d9611c8/Data/LAN2xx/combined_vcf/test_thresholds/"
sample_dir = "/media/mahajrod/d9e6e5ee-1bf7-4dba-934e-3f898d9611c8/Data/LAN2xx/combined_vcf/SNP_annotated_raw_vcf/"
for sample_set in sample_set_names_list:
print("Handling %s" % sample_set)
os.chdir(workdir)
os.system("mkdir -p %s" % sample_set)
os.chdir(sample_set)
mutations = CollectionVCF(vcf_file=sample_dir + sample_set + suffix,
from_file=True)
print("Totaly %s mutations" % len(mutations))
mutations.test_thresholds(extracting_method='distance',
threshold=(50, 5000, 100),
testing_dir="testing_threshold")
def homogeneity_plot(sample_set_names_list, plot_file_prefix):
rcParams.update({'font.size': 6})
letter_list_part1 = ["A", "B", "C", "D", "E", "F", "G", "H", "I", "J"]
feature_type_list = ["all", "pre_5'_UTR", "5'_UTR", "CDS", "3'_UTR"]
figure = plt.figure(1,
dpi=600,
figsize=(5 * 1.47, 2 * len(sample_set_names_list)))
index = 1
for sample, letter in zip(sample_set_names_list, letter_list_part1):
findex = 1
for feature_type in feature_type_list:
vcf_file = "%s_good.vcf" % sample if feature_type == "all" \
else "%s_pre_UTR_variants_only_intergenic_l_300.vcf" % sample \
if feature_type == "pre_5'_UTR" \
else "%s_%s_variants.vcf" % (sample, feature_type)
collection = CollectionVCF(from_file=True, vcf_file=vcf_file)
sample_data = collection.count_strandness("%s_%s_variants" %
(sample, feature_type))
ax = plt.subplot(len(sample_set_names_list),
len(feature_type_list), index)
n_groups = 4
points = np.arange(n_groups)
bar_width = 0.35
C_values = sample_data["all"][0]
G_values = sample_data["all"][1]
rects1 = plt.bar(points,
C_values,
bar_width,
color='b',
label='C->T')
rects2 = plt.bar(points + bar_width,
G_values,
bar_width,
color='g',
label='G->A')
table = [C_values[1:-1], G_values[1:-1]]
g, p_value, dof, expctd = chi2_contingency(table)
phi = phi_coefficient_correlation(table)
if index > len(feature_type_list) * (len(sample_set_names_list) -
1):
plt.xlabel('Strand')
if findex == 1:
plt.ylabel('N of SNV')
plt.text(-0.50,
0.5,
sample[:-3],
rotation=90,
fontweight="bold",
transform=ax.transAxes,
fontsize=10,
horizontalalignment='center',
verticalalignment='center')
#plt.title("%s%i. %s (%i SNV)\np=%.3f, phi=%.3f" % (letter, findex, sample, len(collection), p_value, phi),
# fontweight='bold')
#plt.title("%s%i. %s (%s)\np=%.3f, phi=%.3f" % (letter, findex, sample, feature_type, p_value, phi),
# fontweight='bold', fontsize=6)
title = "%s%i" % (letter, findex)
title_text = r"$p=%.2f, \varphi=%.2f$" % (p_value, phi) if p_value >= 0.01 \
else r"$p=%.1e, \varphi=%.2f$" % (p_value, phi)
plt.text(
0.23,
1.1,
title_text,
rotation=0,
transform=ax.transAxes,
fontsize=8,
#horizontalalignment='center',
verticalalignment='center')
plt.title(title, fontweight='bold', fontsize=11, loc="left")
plt.xticks(points + bar_width, ('None', '+', '-', 'Both'))
if findex == len(feature_type_list):
plt.legend(prop={'size': 8})
if index <= len(feature_type_list):
plt.text(0.5,
1.25,
feature_type,
rotation=0,
fontweight="bold",
transform=ax.transAxes,
fontsize=10,
horizontalalignment='center',
verticalalignment='center')
#plt.suptitle("Strandness histograms", fontweight="bold", fontsize=20)
findex += 1
index += 1
#plt.tight_layout()
plt.subplots_adjust(hspace=0.5,
wspace=0.25,
top=0.88,
left=0.09,
right=0.99)
for extension in [".pdf", ".svg", ".eps", ".png"]:
plt.savefig("%s%s" % (plot_file_prefix, extension))
plt.close()
def read(self, input_file):
# TODO: write read from ccf file
with open(input_file, "r") as in_fd:
stripped_line = in_fd.readline().strip()
if stripped_line == "#VCF_METADATA START":
vcf_metadata = MetadataVCF()
stripped_line = in_fd.readline().strip()
while (stripped_line != "#VCF_METADATA END"):
vcf_metadata.add_metadata(stripped_line)
stripped_line = in_fd.readline().strip()
stripped_line = in_fd.readline().strip()
if stripped_line == "#VCF_HEADER":
header_line = in_fd.readline().strip()
vcf_header = HeaderVCF(header_line[1:].split("\t"))
#print("a\na\na\na\na\n")
#print(vcf_header)
self.metadata = MetadataCCF(vcf_header[9:],
vcf_metadata=vcf_metadata,
vcf_header=vcf_header)
stripped_line = in_fd.readline().strip()
if stripped_line == "#CCF_HEADER":
header_line = in_fd.readline().strip()
self.header = HeaderCCF(header_line[1:].split("\t"))
flag = 0
self.records = []
while True:
data_line = in_fd.readline()
if data_line == "" or data_line == "\n":
break
stripped_line = data_line.strip()
if data_line[0] == "\t":
#stripped_line = stripped_line[1:]
#print(collection_vcf)
collection_vcf.records.append(
collection_vcf.add_record(
stripped_line,
external_metadata=self.metadata.vcf_metadata))
flag = 1
#print("aaaa")
continue
if flag != 0:
self.records.append(
RecordCCF(id=cluster_id,
chrom=chrom,
size=size,
start=start,
end=end,
description=description,
flags=flags,
collection_vcf=collection_vcf,
bad_vcf_records=bad_records,
from_records=False,
subclusters=subclusters))
#collection_vcf = None
if stripped_line[0] == ">":
flag = 0
cluster_id, chrom, start, end, description_and_flags = stripped_line[
1:].split("\t")
start = int(start)
end = int(end)
description_and_flags = description_and_flags.split(";")
description = OrderedDict({})
flags = set([])
subclusters = None
for descr_entry in description_and_flags:
descr_entry_splited = descr_entry.split("=")
if len(descr_entry_splited) == 1:
flags.add(descr_entry_splited[0])
continue
if descr_entry_splited[0] == "Size":
size = int(descr_entry_splited[1])
elif descr_entry_splited[0] == "Bad_records":
bad_records = int(descr_entry_splited[1])
elif descr_entry_splited[
0] == "Mean" or descr_entry_splited[
0] == "Median" or descr_entry_splited[
0] == "Power" or descr_entry_splited[
0] == "Homogeneity":
description[descr_entry_splited[0]] = float(
descr_entry_splited[1])
elif descr_entry_splited[0] == "Loc":
description[descr_entry_splited[
0]] = descr_entry_splited[1].split(",")
elif descr_entry_splited[0] == "Subclusters":
subclusters = [
int(x)
for x in descr_entry_splited[1].split(",")
]
else:
description[descr_entry_splited[
0]] = descr_entry_splited[1].split(",")
if len(description[descr_entry_splited[0]]) == 1:
description[
descr_entry_splited[0]] = description[
descr_entry_splited[0]][0]
collection_vcf = CollectionVCF(metadata=None,
record_list=None,
header=None,
vcf_file=None,
samples=None,
from_file=False,
external_metadata=None)
continue
self.records.append(
RecordCCF(id=cluster_id,
chrom=chrom,
size=size,
start=start,
end=end,
description=description,
flags=flags,
collection_vcf=collection_vcf,
bad_vcf_records=bad_records,
from_records=False,
subclusters=subclusters))
annotations_dict = {}
annotation_synonym_dict = {"three_prime_UTR": "3'_UTR",
"five_prime_UTR": "5'_UTR",
"snoRNA": "ncRNA",
"snRNA": "ncRNA"
}
with open(gff_file) as gff_fd:
for record in GFF.parse(gff_fd):
annotations_dict[record.id] = record
rcParams.update({'font.size': 7})
plt.figure(1, dpi=300, figsize=(8, 6))
index = 1
for sample, letter in zip(sample_set_names_list, letter_list_part1):
collection = CollectionVCF(from_file=True, vcf_file=sample + "_good.vcf")
plt.subplot(3, 2, index)
collection.get_location(annotations_dict, use_synonym=True, synonym_dict=annotation_synonym_dict)
location_pie(collection, annotation_colors=[],
ref_genome=None, explode=True, annotation_black_list=annotation_black_list,
allow_several_counts_of_record=False,
counts_filename="location_counts.t",
counts_dir="location_counts",
legend_font=6,
combine_mixed=True
)
plt.title("%s. %s" % (letter, sample), fontweight='bold')
index += 1
for format_ext in ["svg", "eps", "pdf", "png"]:
vcf_best_merged_homo = "%s_GATK_best_merged_homo.vcf" % sample_name
UnifiedGenotyper.variant_call("%s_trimmed_sorted_rm_pcr_chrom_with_header.bam" % sample_name,
reference,
stand_emit_conf=40,
stand_call_conf=100,
GATK_dir=gatk_dir,
num_of_threads=5,
output_mode="EMIT_VARIANTS_ONLY",
discovery_mode="BOTH",
output_file=vcf_all)
SelectVariants.get_indel(gatk_dir, reference, vcf_all, vcf_indel)
SelectVariants.get_SNP(gatk_dir, reference, vcf_all, vcf_SNP)
VariantFiltration.filter_bad_SNP(gatk_dir, reference, vcf_SNP, vcf_filtered_SNP)
VariantFiltration.filter_bad_indel(gatk_dir, reference, vcf_indel, vcf_filtered_indel)
SelectVariants.remove_filtered(gatk_dir, reference, vcf_filtered_SNP, vcf_best_SNP)
SelectVariants.remove_filtered(gatk_dir, reference, vcf_filtered_indel, vcf_best_indel)
CombineVariants.combine_from_same_source(gatk_dir, reference, [vcf_best_SNP, vcf_best_indel], vcf_best_merged)
best_merged = CollectionVCF(vcf_file=vcf_best_merged)
best_merged_homo, best_merged_hetero = best_merged.split_by_zygoty()
best_merged_homo.write(vcf_best_merged_homo)
best_merged_hetero.write(vcf_best_merged_hetero)
annotations_file = "/home/mahajrod/genetics/desaminases/data/LAN210_v0.10m/annotations/merged_annotations_Nagalakshmi_tranf_to_LAN210_v0.10m.gff3"
sequence_file = "/home/mahajrod/genetics/desaminases/data/LAN210_v0.10m/LAN210_v0.10m.fasta"
workdir = "/media/mahajrod/d9e6e5ee-1bf7-4dba-934e-3f898d9611c8/Data/LAN2xx/combined_vcf"
sequence_dict = SeqIO.to_dict(SeqIO.parse(sequence_file, "fasta"))
#print(sequence_dict)
with open(annotations_file, "r") as in_fd:
annotation_dict = dict([(record.id, record)
for record in GFF.parse(in_fd)])
mutation_strand_dict = {"C": "P", "G": "M"}
#values_names = {"Len": 0, "P": 1, "P_dens": 2, "M": 3, "M_dens": 4, "Exp": 5}
values_names = {"Len": 0, "P": 1, "M": 2, "Exp": 3}
for sample_set_name in sample_set_names_list:
print("Handling %s" % sample_set_name)
os.chdir(workdir)
os.chdir(sample_set_name)
mutations = CollectionVCF(
vcf_file="./clustering/%s_adjusted_cluster_mutations.vcf" %
sample_set_name,
from_file=True)
mutations_large_clusters = CollectionVCF(
vcf_file="./clustering/%s_adjusted_3+_cluster_mutations.vcf" %
sample_set_name,
from_file=True)
prepare_data(mutations, "all_adjusted_cluster_mutations",
expression_data_dir)
prepare_data(mutations_large_clusters, "adjusted_3+_cluster_mutations",
expression_data_dir)
workdir = "/media/mahajrod/d9e6e5ee-1bf7-4dba-934e-3f898d9611c8/Data/LAN2xx/combined_vcf/clusters/all/pre_UTR_strandness/"
letter_list_part1 = ["A", "B", "C", "D", "E", "F", "G", "H", "I", "J"]
os.chdir(workdir)
sample_set_names_list = [
"PmCDA1_3d", "PmCDA1_sub1_3d", "PmCDA1_6d", "PmCDA1_sub1_6d"
]
rcParams.update({'font.size': 7})
feature_type_list = ["5'_UTR", "CDS", "3'_UTR"]
for feature_type in feature_type_list:
plt.figure(1, dpi=300, figsize=(6, 6))
index = 1
for sample, letter in zip(sample_set_names_list, letter_list_part1):
collection = CollectionVCF(from_file=True,
vcf_file="%s_%s_variants.vcf" %
(sample, feature_type))
sample_data = collection.count_strandness("%s_%s_variants" %
(sample, feature_type))
plt.subplot(2, 2, index)
n_groups = 4
points = np.arange(n_groups)
bar_width = 0.35
C_values = sample_data["all"][0]
G_values = sample_data["all"][1]
rects1 = plt.bar(points,
C_values,
bar_width,
color='b',
