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 check_create_metagene(test, values):
metagene = Metagene(*values)
length = sum(values)
(interval, upstream, downstream) = values
expected = "Upstream:{} -- Interval:{} -- Downstream:{}\tLength:{}".format(upstream, interval, downstream, length)
test_description = "\nTest: \t{}\n".format(test)
test_description += "Expected:\t{}\n".format(expected)
test_description += "Metagene:\t{}\n".format(metagene)
assert str(metagene) == expected, "{}Error: \tMetagene does not match expected.".format(test_description)
def setup():
"""Create fixtures"""
# Define chromosome sizes
Read.extract_chromosome_sizes([
"@HD\tVN:1.0\tSO:unsorted", "@SQ\tSN:chr1\tLN:300",
"@SQ\tSN:chr2\tLN:200", "@PG\tID:test\tVN:0.1"
])
Feature.process_set_chromosome_conversion(["1\tchr1", "2\tchr2"])
good_input["bed input counting all of the read"] = (
"all",
"[17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42]"
)
good_input["bed input counting start of the read"] = (
"start", "[17, 18, 19, 20, 21, 22, 23]")
good_input["bed input counting end of the read"] = (
"end", "[36, 37, 38, 39, 40, 41, 42]")
good_input["gff input counting all of the read"] = (
"all",
"[43, 42, 41, 40, 39, 38, 37, 36, 35, 34, 33, 32, 31, 30, 29, 28, 27, 26, 25, 24, 23, 22, 21, 20, 19, 18, 17, 16, 15, 14, 13, 12, 11, 10, 9, 8]"
)
good_input["gff input counting start of the read"] = (
"start", "[43, 42, 41, 40, 39, 38, 37]")
good_input["gff input counting end of the read"] = (
"end", "[14, 13, 12, 11, 10, 9, 8]")
for method in ['all', 'start', 'end']:
print "\nTesting feature_count option: ****{}****".format(method)
if method == 'all':
metagene = Metagene(10, 4, 2)
print "\t with Metagene:\t{}".format(metagene)
print "\t with chromosome conversions:\t{}".format(
Feature.chromosome_conversion)
else:
metagene = Metagene(1, 4, 2)
print "\t with Metagene:\t{}".format(metagene)
print "\t with chromosome conversions:\t{}".format(
Feature.chromosome_conversion)
# create feature from BED line
try:
bedline = "{}\t{}\t{}\t{}\t{}\t{}\n".format(
1, 20, 40, "first", 44, "+")
print "\t with BED line:\t{}".format(bedline.strip())
feature1 = Feature.create_from_bed(method, metagene, bedline,
False, False)
if str(feature1.position_array) != correct_features['bed'][method]:
print "**FAILED**\t Create Feature from BED line ?"
print "\t Desired positions:\t{}".format(
correct_features['bed'][method])
print "\t Created positions:\t{}".format(
feature1.position_array)
except MetageneError as err:
print "**FAILED**\t Create Feature from BED line ?"
else:
print "PASSED\t Create Feature from BED line ?\t\t{}".format(
feature1.get_chromosome_region())
# create feature from GFF line
try:
gffline = "{}\t{}\t{}\t{}\t{}\t{}\t{}\t{}\t{}\n".format(
2, "test", "gene", 10, 39, ".", "-", ".", "second")
print "\t with GFF line:\t{}".format(gffline.strip())
feature2 = Feature.create_from_gff(method, metagene, gffline,
False, False)
if str(feature2.position_array) != correct_features['gff'][method]:
print "**FAILED**\t Create Feature from GFF line ?\t**FAIL**"
print "\t Desired positions:\t{}".format(
correct_features['gff'][method])
print "\t Created positions:\t{}".format(
feature2.position_array)
except MetageneError as err:
print "**FAILED**\t Create Feature from GFF line ?"
else:
print "PASSED\t Create Feature from GFF line ?\t\t{}".format(
feature2.get_chromosome_region())
# create feature from GFF line with start and end swapped
try:
gffline = "{}\t{}\t{}\t{}\t{}\t{}\t{}\t{}\t{}\n".format(
2, "test", "gene", 39, 10, ".", "-", ".", "second")
print "\t with GFF line:\t{}".format(gffline.strip())
feature2 = Feature.create_from_gff(method, metagene, gffline,
False, False)
if str(feature2.position_array) != correct_features['gff'][method]:
print "**FAILED**\t Create Feature from GFF line with swapped start and end ?\t**FAIL**"
print "\t Desired positions:\t{}".format(
correct_features['gff'][method])
print "\t Created positions:\t{}".format(
feature2.position_array)
except MetageneError as err:
print "**FAILED**\t Create Feature from GFF line with swapped start and end ?"
else:
print "PASSED\t Create Feature from GFF line with swapped start and end ?\t\t{}".format(
feature2.get_chromosome_region())
try:
gffline = "{}\t{}\t{}\t{}\t{}\t{}\t{}\t{}\t{}\n".format(
2, "test", "gene", 39, 10, ".", "+", ".", "second")
print "\t with GFF line:\t{}".format(gffline.strip())
feature2 = Feature.create_from_gff(method, metagene, gffline,
False, False)
if str(feature2.position_array) != correct_features['gff'][method]:
print "**FAILED**\t Do not create Feature from GFF line with swapped start and end, + strand ?\t**FAIL**"
print "\t Desired positions:\t{}".format(
correct_features['gff'][method])
print "\t Created positions:\t{}".format(
feature2.position_array)
except MetageneError as err:
print "PASSED\t Do not create Feature from GFF line with swapped start and end, + strand ?\t\t{}".format(
err)
else:
print "**FAILED**\t Do not create Feature from GFF line with swapped start and end, + strand ?\t\t{}".format(
feature2.get_chromosome_region())
##TODO finish complete testing of Feature class
print "\n##TODO finish testing of Feature class creation\n"
print "\n**** Testing counting and maniputlation ****\n"
expected = {'all': {}, 'start': {}, 'end': {}}
# Positions in metagene: 17 18 19 20 21-22,23-24,25-26,27-28,29-30,31-32,33-34,35-36,37-38,39-40, 41, 42
expected['all'] = {
'all':
"first,sense:allreads,0.333,0.333,0.000,0.000,0.000,0.000,0.000,0.000,0.286,0.571,0.571,0.000,0.000,0.286,0.286,0.000\nfirst,antisense:allreads,0.000,0.000,0.000,0.000,0.000,0.100,0.100,0.100,0.100,0.100,0.000,0.000,0.000,0.000,0.000,0.111",
'start':
"first,sense:allreads,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,2.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000\nfirst,antisense:allreads,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.500,0.000,0.000,0.000,0.000,0.000,0.000",
'end':
"first,sense:allreads,0.000,3.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,2.000,0.000\nfirst,antisense:allreads,0.000,0.000,0.000,0.000,0.000,0.500,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,1.000"
}
# Positions in metagene: 17 18 19 20 [21] 22 23
expected['start'] = {
'all':
"first,sense:allreads,0.333,0.333,0.000,0.000,0.000,0.000,0.000\nfirst,antisense:allreads,0.000,0.000,0.000,0.000,0.000,0.000,0.050",
'start':
"first,sense:allreads,0.000,0.000,0.000,0.000,0.000,0.000,0.000\nfirst,antisense:allreads,0.000,0.000,0.000,0.000,0.000,0.000,0.000",
'end':
"first,sense:allreads,0.000,3.000,0.000,0.000,0.000,0.000,0.000\nfirst,antisense:allreads,0.000,0.000,0.000,0.000,0.000,0.000,0.500"
}
# Positions in metagene: 36 37 38 39 [40] 41 42
expected['end'] = {
'all':
"first,sense:allreads,0.000,0.000,0.000,0.000,0.286,0.286,0.000\nfirst,antisense:allreads,0.000,0.000,0.000,0.000,0.000,0.000,0.111",
'start':
"first,sense:allreads,0.000,0.000,0.000,0.000,0.000,0.000,0.000\nfirst,antisense:allreads,0.000,0.000,0.000,0.000,0.000,0.000,0.000",
'end':
"first,sense:allreads,0.000,0.000,0.000,0.000,0.000,2.000,0.000\nfirst,antisense:allreads,0.000,0.000,0.000,0.000,0.000,0.000,1.000"
}
metagene = {
'all': Metagene(10, 4, 2),
'start': Metagene(1, 4, 2),
'end': Metagene(1, 4, 2)
}
for method in ['all', 'start', 'end']:
if method == 'all':
print "\t with Metagene:\t{}".format(metagene[method])
print "\t with chromosome conversions:\t{}".format(
Feature.chromosome_conversion)
else:
print "\t with Metagene:\t{}".format(metagene[method])
print "\t with chromosome conversions:\t{}".format(
Feature.chromosome_conversion)
print "\nTesting feature_count option: ****{}****".format(method)
feature_line = "{}\t{}\t{}\t{}\t{}\t{}\n".format(
1, 20, 40, "first", 44, "+")
feature1 = Feature.create_from_bed(method, metagene[method],
feature_line, False, False)
print "\tFeature:\t{}".format(feature1.position_array)
reads = []
reads.append(
Read("chr1", "+", 3, 1, [10, 11, 12, 13, 14, 15, 16, 17, 18]))
reads.append(
Read("chr1", "-", 1, 2, [23, 24, 25, 26, 27, 28, 29, 30, 31, 32]))
reads.append(Read("chr1", "+", 4, 2, [30, 31, 32, 33, 34, 40, 41]))
reads.append(
Read("chr1", "-", 1, 1, [42, 43, 44, 45, 46, 47, 48, 49, 50]))
reads.append(Read("chr1", "+", 10, 1, [51, 52, 53, 54, 55]))
reads.append(Read("chr2", "+", 10, 1,
[18, 19, 20, 21, 22, 23, 24, 25]))
# starting count
for count_method in ['all', 'start', 'end']:
print "\nTesting count_method option: ****{}****".format(
count_method)
output = "{}\n".format(feature1)
for r in reads:
output += "{}\n".format(r)
feature1.count_read(r, count_method, count_partial_reads=True)
output += "{}\n".format(feature1)
output += feature1.print_metagene(pretty=True)
if str(feature1.print_metagene()).strip() == str(
expected[method][count_method]).strip():
print "PASSED\tCreated correct metagene with feature method {} and count method {} ?".format(
method, count_method)
else:
print "**FAILED**\tCreated correct metagene with feature method {} and count method {} ?".format(
method, count_method)
print "\tExpected:\n{}".format(expected[method][count_method])
print "\tActual :\n{}".format(feature1.print_metagene())
print "\tSummary of run:\n{}".format(output)
feature1 = Feature.create_from_bed(
method, metagene[method], feature_line, False,
False) # zero out counter for next round
try:
unstranded_read = Read("chr1", ".", 10, 1,
[18, 19, 20, 21, 22, 23, 24, 25])
feature1.count_read(unstranded_read, 'all')
except MetageneError as err:
print "PASSED\tCaught unstranded read on stranded count ?\t\t".format(
err)
else:
print "**FAILED**\tCaught unstranded read on stranded count ?"
try:
feature_line = "{}\t{}\t{}\t{}\t{}\t{}\n".format(
1, 20, 40, "first", 44, ".")
feature1 = Feature.create_from_bed(method, metagene[method],
feature_line, False, False)
unstranded_read = Read("chr1", ".", 10, 1,
[18, 19, 20, 21, 22, 23, 24, 25])
feature1.count_read(unstranded_read, 'all')
except MetageneError as err:
print "**FAILED**\tAllowed unstranded read on unstranded count ?\t\t".format(
err)
else:
print "PASSED\tAllowed unstranded read on unstranded count ?"
print "\n**** Testing adjust_to_metagene ****\n"
chromosome_converter = {"1": "chr1", "2": "chr2"}
# ((metagene_tupple),(feature_tupple),expected_result_string, message_string)
tests = [((8, 2, 2), (16, 8, 24, 4),
'8.000,8.000,4.000,4.000,12.000,12.000,2.000,2.000',
"Expand to metagene ?"),
((4, 2, 2), (6, 8, 6, 2, 4, 4, 2, 4, 24, 8),
'17.000,9.000,8.000,34.000', "Contract to metagene ?"),
((4, 2, 2), (2.5, 4, (10.0 / 3), 10, 11, 7.3, 4),
'5.500,9.333,17.825,9.475', "Contract with messy floats ?"),
((3, 2, 2), (2.5, 4, (10.0 / 3), 10, 11, 7.3, 4),
'7.611,19.556,14.967', "Contract with other messy floats ?")]
for t in tests:
metagene = Metagene(*t[0])
print "\t{}".format(metagene)
feature_line = "{}\t{}\t{}\n".format(1, 0, len(t[1]))
feature = Feature.create_from_bed('all',
metagene,
feature_line,
False,
False,
short=True)
adjusted_feature = ""
for f in feature.adjust_to_metagene(t[1]):
adjusted_feature += "{0:0.3f},".format(f)
if adjusted_feature[:-1] == t[2]:
print "PASSED\t{}".format(t[3])
else:
print "**FAILED**\t{}".format(t[3])
print "\tExpected:\t{}".format(t[2])
print "\tActual :\t{}".format(adjusted_feature[:-1])
print "\tOriginal:\t{}".format(feature.adjust_to_metagene(t[1]))
print "\n**** End of Testing the Feature class ****\n"
# end of Feature.test method
def check_catch_bad_input(test, values):
print Metagene(*values)
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 __init__(self,
count_method,
metagene_object,
name,
chromosome,
start,
end,
strand,
gap_counting=False,
ignore_strand=False):
"""Not normally called directly; use Feature.create(file_format, count_method,
metagene_object, feature_line, chromosome_conversion_table) to call indirectly.
Define a new feature with an interval (represents feature length),
up and downstream padding (defined by metagene_object), and genomic
(1-based) start and end positions.
Once defined here, the start and end represent the true start and end of
the feature. Therefore, if a - strand (Crick strand) feature the start
will be larger than the end.
"""
chromosome = Feature.chromosome_conversion[
chromosome] # convert to BAM-like chromosome designation
if (confirm_integer(start,
"Start",
minimum=1,
maximum=Read.chromosome_sizes[chromosome]) and
confirm_integer(end,
"End",
minimum=1,
maximum=Read.chromosome_sizes[chromosome])):
start = int(start)
end = int(end)
# Define feature-specific metagene where feature_interval respresents
# the length of the feature NOT the length of the final metagene interval
if count_method == 'all':
interval = (end - start + 1) # length of feature
else:
interval = 1 # length of the start (or end) of feature
Metagene.__init__(self, interval, metagene_object.padding['Upstream'],
metagene_object.padding['Downstream'])
self.name = name
self.chromosome = chromosome
self.strand = strand
self.metagene_length = metagene_object.feature_interval
# define counts_array dictionary
# key: orientation:gap_counts string
# where orientation = {'unstranded', 'sense', 'antisense'}
# gap_counts = {'ungapped', 'gapped, 'allreads'}
# 'ungapped' + 'gapped' = 'allreads'
# 'sense' + 'antisense' = 'unstranded'
#
# values: arrays of self.length initialized to 0
if self.strand != "+" and self.strand != "-":
self.strand = "."
orientation = ['unstranded']
elif ignore_strand:
orientation = ['unstranded']
else:
orientation = ['sense', 'antisense']
if gap_counting:
gap_counts = ['ungapped', 'gapped']
else:
gap_counts = ['allreads']
self.counts_array = {}
for o in orientation:
for g in gap_counts:
self.counts_array["{}:{}".format(o, g)] = []
for p in range(self.length):
#self.counts_array["{}:{}".format(o,g)].append(decimal.Decimal(0.0))
self.counts_array["{}:{}".format(o, g)].append(0)
# define position_array
# values : chromosomal 1-based nucleotide positions in 5' to 3'
# orientation WITH RESPECT TO THE FEATURE
# Example :
# + strand: [10,11,12,13,14,15]
# - strand: [15,14,13,12,11,10]
# so position_array[0] is always the start of the feature (with upstream padding)
# position_array[-1] is always the end of the feature (with downstream padding)
self.position_array = []
if self.strand == "-":
# chromosome start = feature end
# chromosome end = feature start
if count_method == 'start':
start = end
elif count_method == 'end':
end = start
region_start = start - self.padding[
'Downstream'] # start is really end
region_end = end + self.padding['Upstream'] # end is really start
positions = range(region_start, region_end + 1) # inclusive list
positions.reverse()
else:
if count_method == 'start':
end = start # set both start and end to the start value
elif count_method == 'end':
start = end # set both start and end to the end value
region_start = start - self.padding['Upstream']
region_end = end + self.padding['Downstream']
positions = range(region_start, region_end + 1) # inclusive list
self.position_array = positions
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 __init__(
self,
count_method,
metagene_object,
name,
chromosome,
start,
end,
strand,
gap_counting=False,
ignore_strand=False,
):
"""Not normally called directly; use Feature.create(file_format, count_method,
metagene_object, feature_line, chromosome_conversion_table) to call indirectly.
Define a new feature with an interval (represents feature length),
up and downstream padding (defined by metagene_object), and genomic
(1-based) start and end positions.
Once defined here, the start and end represent the true start and end of
the feature. Therefore, if a - strand (Crick strand) feature the start
will be larger than the end.
"""
chromosome = Feature.chromosome_conversion[chromosome] # convert to BAM-like chromosome designation
if confirm_integer(start, "Start", minimum=1, maximum=Read.chromosome_sizes[chromosome]) and confirm_integer(
end, "End", minimum=1, maximum=Read.chromosome_sizes[chromosome]
):
start = int(start)
end = int(end)
# Define feature-specific metagene where feature_interval respresents
# the length of the feature NOT the length of the final metagene interval
if count_method == "all":
interval = end - start + 1 # length of feature
else:
interval = 1 # length of the start (or end) of feature
Metagene.__init__(self, interval, metagene_object.padding["Upstream"], metagene_object.padding["Downstream"])
self.name = name
self.chromosome = chromosome
self.strand = strand
self.metagene_length = metagene_object.feature_interval
# define counts_array dictionary
# key: orientation:gap_counts string
# where orientation = {'unstranded', 'sense', 'antisense'}
# gap_counts = {'ungapped', 'gapped, 'allreads'}
# 'ungapped' + 'gapped' = 'allreads'
# 'sense' + 'antisense' = 'unstranded'
#
# values: arrays of self.length initialized to 0
if self.strand != "+" and self.strand != "-":
self.strand = "."
orientation = ["unstranded"]
elif ignore_strand:
orientation = ["unstranded"]
else:
orientation = ["sense", "antisense"]
if gap_counting:
gap_counts = ["ungapped", "gapped"]
else:
gap_counts = ["allreads"]
self.counts_array = {}
for o in orientation:
for g in gap_counts:
self.counts_array["{}:{}".format(o, g)] = []
for p in range(self.length):
# self.counts_array["{}:{}".format(o,g)].append(decimal.Decimal(0.0))
self.counts_array["{}:{}".format(o, g)].append(0)
# define position_array
# values : chromosomal 1-based nucleotide positions in 5' to 3'
# orientation WITH RESPECT TO THE FEATURE
# Example :
# + strand: [10,11,12,13,14,15]
# - strand: [15,14,13,12,11,10]
# so position_array[0] is always the start of the feature (with upstream padding)
# position_array[-1] is always the end of the feature (with downstream padding)
self.position_array = []
if self.strand == "-":
# chromosome start = feature end
# chromosome end = feature start
if count_method == "start":
start = end
elif count_method == "end":
end = start
region_start = start - self.padding["Downstream"] # start is really end
region_end = end + self.padding["Upstream"] # end is really start
positions = range(region_start, region_end + 1) # inclusive list
positions.reverse()
else:
if count_method == "start":
end = start # set both start and end to the start value
elif count_method == "end":
start = end # set both start and end to the end value
region_start = start - self.padding["Upstream"]
region_end = end + self.padding["Downstream"]
positions = range(region_start, region_end + 1) # inclusive list
self.position_array = positions
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))