def print_metagene(self, pretty=False, header=False, interval_override=False):
"""Converts counts_array data to finalized metagene profiles for printing
Standard printing is in comma-delimited lines for input into metagene_analysis.py
Pretty printing (pretty=True) gives a human readable, if potentially super long, version
"""
final_metagenes = {}
if interval_override:
metagene = Metagene(self.feature_interval, self.padding["Upstream"], self.padding["Downstream"])
output = "# Metagene:\t{}\n".format(metagene)
output += metagene.print_full()
elif header:
metagene = Metagene(self.metagene_length, self.padding["Upstream"], self.padding["Downstream"])
output = metagene.print_full(pretty)
else:
output = ""
# process each subset grouping
for subset in sorted(self.counts_array, reverse=True):
# break counts_array into sections -> upstream padding, interval_feature, and downstream padding
upstream_counts = self.counts_array[subset][0 : self.padding["Upstream"]]
interval_counts = self.counts_array[subset][
self.padding["Upstream"] : self.padding["Upstream"] + self.feature_interval
]
downstream_counts = self.counts_array[subset][
self.padding["Upstream"] + self.feature_interval : len(self.counts_array[subset])
]
if interval_override:
metagene_interval_counts = interval_counts
else:
# compress (or expand) interval_counts to match the size of the internal metagene
metagene_interval_counts = self.adjust_to_metagene(interval_counts)
if pretty:
output += "{0:15s}:\t".format(subset)
for i in upstream_counts:
output += "{0:>5s},".format("{0:3.2f}".format(i)) # keep 2 decimal places in the outputted float
for i in metagene_interval_counts:
output += "{0:>5s},".format("{0:3.2f}".format(i))
for i in downstream_counts:
output += "{0:>5s},".format("{0:3.2f}".format(i))
output = output[:-1] + "\n"
else:
# build output
output += "{},{}".format(self.name, subset)
for p in upstream_counts:
output += ",{0:0.3f}".format(p) # keep 3 decimal places in the outputted float
for p in metagene_interval_counts:
output += ",{0:0.3f}".format(p)
for p in downstream_counts:
output += ",{0:0.3f}".format(p)
output += "\n"
return output.strip() # remove trailing "\n"
def check_print_metagene_plain(test, values):
expected = (-values[1], values[0] + values[2] - 1)
metagene = Metagene(*values)
plain_print = metagene.print_full().strip()
plain_print_parts = plain_print.split(",")
new_range = (int(plain_print_parts[2]), int(plain_print_parts[-1]))
print "\n\tOutput:\n\t{}".format(plain_print)
test_description = "\nTest: \t{}\n".format(test)
test_description += "Expected:\t{}\n".format(expected)
test_description += "Range: \t{}\n".format(new_range)
test_description += "Output: \t{}\n".format(plain_print)
assert new_range == expected, "{}Error: \tPrinted metagene does not match expected.".format(test_description)
def check_print_metagene_pretty(test, values):
metagene = Metagene(*values)
pretty_print = metagene.print_full(pretty=True).strip()
new_values = (len(re.findall('int', pretty_print)),
len(re.findall('up', pretty_print)),
len(re.findall('down', pretty_print)))
print "\n\tOutput:\n\t{}".format("\n\t".join(pretty_print.split("\n")))
test_description = "\nTest: \t{}\n".format(test)
test_description += "Expected:\t{}\n".format(values)
test_description += "Range: \t{}\n".format(new_values)
test_description += "Output: \t{}\n".format(pretty_print)
assert new_values == values, "{}Error: \tPrinted metagene does not match expected.".format(test_description)
def print_metagene(self,
pretty=False,
header=False,
interval_override=False):
"""Converts counts_array data to finalized metagene profiles for printing
Standard printing is in comma-delimited lines for input into metagene_analysis.py
Pretty printing (pretty=True) gives a human readable, if potentially super long, version
"""
final_metagenes = {}
if interval_override:
metagene = Metagene(self.feature_interval,
self.padding['Upstream'],
self.padding['Downstream'])
output = "# Metagene:\t{}\n".format(metagene)
output += metagene.print_full()
elif header:
metagene = Metagene(self.metagene_length, self.padding['Upstream'],
self.padding['Downstream'])
output = metagene.print_full(pretty)
else:
output = ""
# process each subset grouping
for subset in sorted(self.counts_array, reverse=True):
# break counts_array into sections -> upstream padding, interval_feature, and downstream padding
upstream_counts = self.counts_array[subset][0:self.
padding['Upstream']]
interval_counts = self.counts_array[
subset][self.padding['Upstream']:self.padding['Upstream'] +
self.feature_interval]
downstream_counts = self.counts_array[
subset][self.padding['Upstream'] +
self.feature_interval:len(self.counts_array[subset])]
if interval_override:
metagene_interval_counts = interval_counts
else:
# compress (or expand) interval_counts to match the size of the internal metagene
metagene_interval_counts = self.adjust_to_metagene(
interval_counts)
if pretty:
output += "{0:15s}:\t".format(subset)
for i in upstream_counts:
output += "{0:>5s},".format("{0:3.2f}".format(
i)) # keep 2 decimal places in the outputted float
for i in metagene_interval_counts:
output += "{0:>5s},".format("{0:3.2f}".format(i))
for i in downstream_counts:
output += "{0:>5s},".format("{0:3.2f}".format(i))
output = output[:-1] + "\n"
else:
# build output
output += "{},{}".format(self.name, subset)
for p in upstream_counts:
output += ",{0:0.3f}".format(
p) # keep 3 decimal places in the outputted float
for p in metagene_interval_counts:
output += ",{0:0.3f}".format(p)
for p in downstream_counts:
output += ",{0:0.3f}".format(p)
output += "\n"
return output.strip() # remove trailing "\n"
def metagene_count():
"""Chain of command for metagene_count analysis."""
arguments = get_arguments()
# confirm BAM file and extract chromosome sizes
Read.set_chromosome_sizes(arguments.alignment)
##TODO: create a list of chromosomes to analyze and/or exclude
# create chromosome conversion dictionary for feature (GFF/BED) to alignment (BAM)
Feature.set_chromosome_conversion(arguments.chromosome_names, Read.chromosome_sizes.keys())
# define has_abundance and has_mappings tags for Read class
Read.set_sam_tag(arguments.extract_abundance, arguments.alignment, "NA:i:(\d+)")
Read.set_sam_tag(arguments.extract_mappings, arguments.alignment, "NH:i:(\d+)")
# define the metagene array shape (left padding, start, internal, end, right padding)
# metagene = padding ---- internal region ---- padding
try:
metagene = Metagene(arguments.interval_size, arguments.padding, arguments.padding)
print "Metagene definition:\t{}".format(metagene)
except MetageneError as err:
print err
raise MetageneError("Unable to create the metagene template")
try:
Feature.set_format(arguments.feature) # assign file format for the feature file
print "Reading feature file as {} format".format(Feature.format)
except MetageneError as err:
print err
raise MetageneError("Unable to create the feature object")
# print out the header line...
if not arguments.interval_variable:
with open("{}.metagene_counts.csv".format(arguments.output_prefix), 'w') as output_file:
output_file.write("# Metagene:\t{}\n".format(metagene)) # define for plotting later
output_file.write(metagene.print_full())
# for each feature
with open(arguments.feature, 'r') as feature_file:
for feature_line in read_chunk(feature_file, 1024):
if feature_line[0] != "#": # skip comment lines
# change creation with feature_method
feature = Feature.create(arguments.feature_count, metagene, feature_line, arguments.count_splicing,
arguments.ignore_strand)
# pull out sam file lines; it is important to use Feature.get_samtools_region(chromosome_lengths) rather
# than Feature.get_chromosome_region() because only the first ensures that the interval does not
# extend beyond the length of the chromosome which makes samtools view return no reads
(run_pipe_worked, sam_sample) = run_pipe(['samtools view {} {}'.format(
arguments.alignment,
feature.get_samtools_region())])
if run_pipe_worked:
for samline in sam_sample:
if len(samline) > 0:
# create Read feature
(created_read, read) = Read.create_from_sam(samline,
Feature.chromosome_conversion.values(),
arguments.count_method,
arguments.uniquely_mapping,
arguments.ignore_strand,
arguments.count_secondary_alignments,
arguments.count_failed_quality_control,
arguments.count_PCR_optical_duplicate,
arguments.count_supplementary_alignment)
# count read (if it exists)
if created_read:
feature.count_read(read, arguments.count_method, arguments.count_splicing,
arguments.count_partial_reads, arguments.ignore_strand)
# output the resulting metagene
with open("{}.metagene_counts.csv".format(arguments.output_prefix), 'a') as output_file:
output_file.write(
"{}\n".format(feature.print_metagene(interval_override=arguments.interval_variable)))
else:
raise MetageneError("Could not pull chromosomal region {} for feature {} from BAM file {}.".format(
feature.get_chromosome_region(),
feature.name,
arguments.alignment))
def metagene_count():
"""Chain of command for metagene_count analysis."""
arguments = get_arguments()
# confirm BAM file and extract chromosome sizes
Read.set_chromosome_sizes(arguments.alignment)
##TODO: create a list of chromosomes to analyze and/or exclude
# create chromosome conversion dictionary for feature (GFF/BED) to alignment (BAM)
Feature.set_chromosome_conversion(arguments.chromosome_names,
Read.chromosome_sizes.keys())
# define has_abundance and has_mappings tags for Read class
Read.set_sam_tag(arguments.extract_abundance, arguments.alignment,
"NA:i:(\d+)")
Read.set_sam_tag(arguments.extract_mappings, arguments.alignment,
"NH:i:(\d+)")
# define the metagene array shape (left padding, start, internal, end, right padding)
# metagene = padding ---- internal region ---- padding
try:
metagene = Metagene(arguments.interval_size, arguments.padding,
arguments.padding)
print "Metagene definition:\t{}".format(metagene)
except MetageneError as err:
print err
raise MetageneError("Unable to create the metagene template")
try:
Feature.set_format(
arguments.feature) # assign file format for the feature file
print "Reading feature file as {} format".format(Feature.format)
except MetageneError as err:
print err
raise MetageneError("Unable to create the feature object")
# print out the header line...
if not arguments.interval_variable:
with open("{}.metagene_counts.csv".format(arguments.output_prefix),
'w') as output_file:
output_file.write("# Metagene:\t{}\n".format(
metagene)) # define for plotting later
output_file.write(metagene.print_full())
# for each feature
with open(arguments.feature, 'r') as feature_file:
for feature_line in read_chunk(feature_file, 1024):
if feature_line[0] != "#": # skip comment lines
# change creation with feature_method
feature = Feature.create(arguments.feature_count, metagene,
feature_line,
arguments.count_splicing,
arguments.ignore_strand)
# pull out sam file lines; it is important to use Feature.get_samtools_region(chromosome_lengths) rather
# than Feature.get_chromosome_region() because only the first ensures that the interval does not
# extend beyond the length of the chromosome which makes samtools view return no reads
(run_pipe_worked, sam_sample) = run_pipe([
'samtools view {} {}'.format(arguments.alignment,
feature.get_samtools_region())
])
if run_pipe_worked:
for samline in sam_sample:
if len(samline) > 0:
# create Read feature
(created_read, read) = Read.create_from_sam(
samline,
Feature.chromosome_conversion.values(),
arguments.count_method,
arguments.uniquely_mapping,
arguments.ignore_strand,
arguments.count_secondary_alignments,
arguments.count_failed_quality_control,
arguments.count_PCR_optical_duplicate,
arguments.count_supplementary_alignment)
# count read (if it exists)
if created_read:
feature.count_read(
read, arguments.count_method,
arguments.count_splicing,
arguments.count_partial_reads,
arguments.ignore_strand)
# output the resulting metagene
with open(
"{}.metagene_counts.csv".format(
arguments.output_prefix), 'a') as output_file:
output_file.write("{}\n".format(
feature.print_metagene(interval_override=arguments.
interval_variable)))
else:
raise MetageneError(
"Could not pull chromosomal region {} for feature {} from BAM file {}."
.format(feature.get_chromosome_region(), feature.name,
arguments.alignment))