def test_can_find_chrom_using_legacy_argument_name(self):
line = '1\t100\tA\tC\t1'
special_parser = parsers.GenericGwasLineParser(chr_col=1, pos_col=2, ref_col=3, alt_col=4,
pvalue_col=5, is_neg_log_pvalue=True,
delimiter='\t')
p = special_parser(line)
assert p.chrom == '1'
def standard_gwas_parser_basic():
return parsers.GenericGwasLineParser(chrom_col=1,
pos_col=2,
ref_col=3,
alt_col=4,
pvalue_col=5,
is_neg_log_pvalue=True,
delimiter='\t')
def test_can_convert_to_neglogpvalue(self):
line = '1\t100\tA\tC\t1'
special_parser = parsers.GenericGwasLineParser(chrom_col=1, pos_col=2, ref_col=3, alt_col=4,
pvalue_col=5, is_neg_log_pvalue=True,
delimiter='\t')
p = special_parser(line)
assert p.neg_log_pvalue == pytest.approx(1), 'Converts -log to pvalue'
assert p.pvalue == pytest.approx(0.1), 'Converts -log to pvalue'
def test_can_convert_to_logpvalue_using_legacy_argument_names(self):
line = '1\t100\tA\tC\t1'
special_parser = parsers.GenericGwasLineParser(chrom_col=1, pos_col=2, ref_col=3, alt_col=4,
pval_col=5, is_log_pval=True,
delimiter='\t')
p = special_parser(line)
assert p.neg_log_pvalue == pytest.approx(1), 'Parses -logp as is'
assert p.pvalue == pytest.approx(0.1), 'Converts -log to pvalue'
def test_gets_marker_info_from_hybrid_fields(self):
line = 'chr2:100_NA_NA\tA\tC\t.05'
special_parser = parsers.GenericGwasLineParser(marker_col=1, ref_col=2, alt_col=3, pval_col=4)
p = special_parser(line)
assert p.chrom == '2', 'Read chrom from marker'
assert p.pos == 100, 'Read pos from marker'
assert p.ref == 'A', 'Read ref from column and ignored marker value'
assert p.alt == 'C', 'Read alt from column and ignored marker value'
def test_warns_about_incorrect_delimiter(self):
"""
Regression test: human-edited files may have a mix of tabs and spaces; this is hard to spot!
"""
line = 'chr2:100:A:C_anno\t.05'
special_parser = parsers.GenericGwasLineParser(marker_col=1, pvalue_col=2, delimiter=' ')
with pytest.raises(exceptions.LineParseException, match="delimiter"):
special_parser(line)
def test_parses_marker_to_clean_format(self):
line = 'chr2:100:A:C_anno\t.05'
special_parser = parsers.GenericGwasLineParser(marker_col=1, pvalue_col=2, delimiter='\t')
p = special_parser(line)
assert p.chrom == '2', 'Finds chromosome'
assert p.pos == 100, 'Finds position'
assert p.ref == 'A', 'Finds ref'
assert p.alt == 'C', 'Finds alt'
assert p.marker == '2:100_A/C', 'Turns a messy marker into a cleaned standardized format'
def test_parses_rsid_to_clean_format(self):
scenarios = [
('chrx\t100\t.05\trs12', 'rs12'), # Handles valid rsid as given
('chrx\t100\t.05\tNA', None), # Missing values
('chrx\t100\t.05\t99', 'rs99'), # Ensures prefix is present
]
parser = parsers.GenericGwasLineParser(chrom_col=1, pos_col=2, pvalue_col=3, rsid_col=4, delimiter='\t')
for text, expected in scenarios:
actual = parser(text).rsid
assert actual == expected, 'Found correct rsid from: {}'.format(text)
def standard_gwas_parser():
return parsers.GenericGwasLineParser(chrom_col=1,
pos_col=2,
ref_col=3,
alt_col=4,
pvalue_col=5,
is_neg_log_pvalue=True,
beta_col=6,
stderr_beta_col=7,
allele_freq_col=8,
is_alt_effect=True,
delimiter='\t')
def main(source: ty.Union[str, ty.Iterable],
out_fn: ty.Union[str, None],
parser_options: dict,
auto_config=False,
skip_rows=None,
skip_errors=True,
max_errors=100,
make_tabix: bool = False):
try:
parser = parsers.GenericGwasLineParser(**parser_options)
except exceptions.ConfigurationException:
parser = None
if source is None:
source = sys.stdin
if not auto_config and (skip_rows is None or parser is None):
logger.error(
'Please provide all options required to parse the file, or use the --auto flag to guess'
)
sys.exit(1)
# Guess how to read the file. If no parser was provided, try to guess columns.
reader = sniffers.guess_gwas_generic(source,
skip_rows=skip_rows,
parser=parser,
parser_options=parser_options,
skip_errors=skip_errors,
max_errors=max_errors)
try:
dest_fn = reader.write(out_fn, make_tabix=make_tabix) or 'console'
except exceptions.TooManyBadLinesException:
logger.error('ERROR: Too many lines failed to parse; stopping.')
except Exception:
logger.exception('Conversion failed due to unknown error')
else:
logger.info(
'Conversion succeeded! Results written to: {}'.format(dest_fn))
finally:
for n, reason, _ in reader.errors:
logger.error(
'Excluded row {} from output due to parse error: {}'.format(
n, reason))
def test_validates_frequency_fields(self):
with pytest.raises(exceptions.ConfigurationException, match='mutually exclusive'):
parsers.GenericGwasLineParser(marker_col=1, pvalue_col=2, allele_count_col=3, allele_freq_col=4)
with pytest.raises(exceptions.ConfigurationException, match='n_samples'):
parsers.GenericGwasLineParser(marker_col=1, pvalue_col=2, allele_count_col=3, n_samples_col=None)
def test_validates_that_allele_spec_is_none_or_both(self):
with pytest.raises(exceptions.ConfigurationException, match='all required'):
parsers.GenericGwasLineParser(marker_col=1, ref_col=3, pvalue_col=None)
def test_validates_arguments_required_fields(self):
with pytest.raises(exceptions.ConfigurationException, match='all required'):
parsers.GenericGwasLineParser(marker_col=1, pvalue_col=None)
def test_parses_freq_from_freq(self):
line = 'chr2:100:A:C_anno\t.05\t0.25'
special_parser = parsers.GenericGwasLineParser(marker_col=1, pvalue_col=2,
allele_freq_col=3, is_alt_effect=True)
p = special_parser(line)
assert p.alt_allele_freq == 0.25, "Parses frequency as is"
def test_parses_freq_from_counts(self):
line = 'chr2:100:A:C_anno\t.05\t25\t100'
special_parser = parsers.GenericGwasLineParser(marker_col=1, pvalue_col=2,
allele_count_col=3, n_samples_col=4, is_alt_effect=False)
p = special_parser(line)
assert p.alt_allele_freq == 0.875, "Calculates frequency from counts and orients to alt allele"
def test_parses_chr_to_clean_format(self):
line = 'chrx\t100\t.05'
special_parser = parsers.GenericGwasLineParser(chrom_col=1, pos_col=2, pvalue_col=3, delimiter='\t')
p = special_parser(line)
assert p.chrom == 'X', 'Strips prefix from chromosome labels and always uses uppercase letters'
