def main(table_in, table_out, pathways, to_classic):
# setup
table = load_table(table_in)
pathway_dict = get_pathway2kos()
# get set of kos from pathways
pathways_kos = set()
for pathway in pathways:
pathways_kos = pathways_kos | pathway_dict[pathway.strip()[-5:]]
# get selected kos
kos_to_keep = set(table.ids('observation')) & \
pathways_kos
if len(kos_to_keep) == 0:
raise EmptySetERROR('Intersection created empty set')
obs_ids = np.array(list(kos_to_keep))
data = np.empty([len(obs_ids), len(table.ids('sample'))])
for i, obs in enumerate(obs_ids):
data[i] = table.data(obs, 'observation')
# output
new_table = Table(data, obs_ids, table.ids('sample'), type="OTU table")
if to_classic:
# print to tab delimited biom table
f = open(table_out, 'w')
f.write(new_table.to_tsv())
else:
# print json biom table
new_table.to_json("filter_KOs_by_pathway.py", open(table_out, 'w'))
def differentialtest(table: biom.Table, metadata: qiime2.Metadata,
variable: str,
taxonomy: TSVTaxonomyFormat) -> pd.DataFrame:
if table.is_empty():
raise ValueError("The provided table object is empty")
## run the R script on the file
with tempfile.TemporaryDirectory() as temp_dir_name:
## write the biom table to file
input_table = os.path.join(temp_dir_name, 'table.tsv')
input_metadata = os.path.join(temp_dir_name, 'metadata.tsv')
with open(input_table, 'w') as fh:
fh.write(table.to_tsv())
metadata.save(input_metadata)
output = os.path.join(temp_dir_name, 'data.tsv')
cmd = [
'differentialtest.R', input_table, input_metadata,
str(taxonomy),
str(variable),
str(output)
]
run_commands([cmd])
data = pd.read_csv(output, sep='\t')
data.index.name = 'Feature ID'
return data
def SRS(table: biom.Table,
c_min: int,
set_seed: bool = True,
seed: int = 1) -> biom.Table:
if table.is_empty():
raise ValueError("The provided table object is empty")
#normalized_table = biom.Table()
## run the R script on the file
with tempfile.TemporaryDirectory() as temp_dir_name:
## write the biom table to file
input_name = os.path.join(temp_dir_name, 'table.tsv')
with open(input_name, 'w') as fh:
fh.write(table.to_tsv())
cmd = ['SRS.R', input_name, str(c_min), str(set_seed), str(seed)]
run_commands([cmd])
norm_table_df = pd.read_csv(input_name, sep='\t')
norm_table_biom = biom.Table(data=norm_table_df.values,
observation_ids=norm_table_df.index,
sample_ids=norm_table_df.columns)
return norm_table_biom
def SRScurve(output_dir: str,
table: biom.Table,
metric: str = 'richness',
step: int = 50,
sample: int = 0,
max_sample_size: int = 0,
rarefy_comparison: bool = False,
rarefy_repeats: int = 10,
rarefy_comparison_legend: bool = False,
srs_color: str = 'black',
rarefy_color: str = 'red',
srs_linetype: str = 'solid',
rarefy_linetype: str = 'longdash',
label: bool = False) -> None:
if table.is_empty():
raise ValueError("The provided table object is empty")
## run the R script on the file
with tempfile.TemporaryDirectory() as temp_dir_name:
## write the biom table to file
input_name = os.path.join(temp_dir_name, 'table.tsv')
#input_name = 'table.tsv'
with open(input_name, 'w') as fh:
fh.write(table.to_tsv())
#table_df = pd.read_csv(input_name, sep='\t')
cmd = [
'SRScurve.R', input_name,
str(metric),
str(step),
str(sample),
str(max_sample_size),
str(rarefy_comparison),
str(rarefy_repeats),
str(rarefy_comparison_legend),
str(srs_color),
str(rarefy_color),
str(srs_linetype),
str(rarefy_linetype),
str(label),
str(output_dir)
]
run_commands([cmd])
plot = os.path.join(output_dir, 'plot.png')
index = os.path.join(output_dir, 'index.html')
with open(index, 'w') as fh:
fh.write(
'<!DOCTYPE html><head></head><body><img src="SRScurve_plot.png" style="max-width: 100vw;max-height: 100vh;object-fit: contain" /></body></html>'
)
def alpha(table: biom.Table) -> AlphaDiversityFormat:
if table.is_empty():
raise ValueError("The provided table object is empty")
output = AlphaDiversityFormat()
## run the R script on the file
with tempfile.TemporaryDirectory() as temp_dir_name:
## write the biom table to file
input_name = os.path.join(temp_dir_name, 'table.tsv')
with open(input_name, 'w') as fh:
fh.write(table.to_tsv())
cmd = ['run_new_richness.R', input_name, str(output)]
run_commands([cmd])
return output
def generate_per_sample_biom(biom_file, limit):
"""Generate per-sample BIOM files
Parameters
----------
biom_file : str
A filepath to a BIOM table
limit : int or None
Limit the number of tables to load
Returns
-------
str
The sample ID
str
The table in BIOM Format v1.0
str
The table in the classic OTU table format
"""
table = load_table(biom_file)
obs_ids = table.ids(axis='observation')
obs_md = table.metadata(axis='observation')
if limit is None:
limit = np.inf
count = 0
for v, sample, _ in table.iter():
if count >= limit:
break
single_sample = Table(v[:, np.newaxis], obs_ids, [sample], obs_md)
single_sample.filter(lambda v_, i, md: v_ > 0, axis='observation')
biomv1 = single_sample.to_json('AG')
biomtxt = single_sample.to_tsv(
header_key='taxonomy',
header_value='taxonomy',
metadata_formatter=lambda x: '; '.join(x))
yield (sample, biomv1, biomtxt)
count += 1
def cross_validate_map(
table: biom.Table,
groups: pd.DataFrame,
phylogeny: NewickFormat,
full_phylogeny: NewickFormat,
) -> (biom.Table, pd.DataFrame):
with tempfile.TemporaryDirectory() as temp_dir_name:
input_table = os.path.join(temp_dir_name, 'table.tsv')
with open(input_table, 'w') as fh:
fh.write(table.to_tsv())
input_groups = os.path.join(temp_dir_name, 'groups.tsv')
groups.to_csv(input_groups, sep='\t')
biom_output = os.path.join(temp_dir_name, 'out_table.tsv')
group_output = os.path.join(temp_dir_name, 'groups_out.tsv')
cmd = [
'run_crossVmap.R', input_table, input_groups,
str(phylogeny),
str(full_phylogeny),
str(biom_output),
str(group_output)
]
try:
print('Running Commands')
run_commands([cmd])
except subprocess.CalledProcessError as e:
raise Exception("An error was encountered with PhyloFactor"
" in R (return code %d), please inspect stdout"
" and stderr to learn more." % e.returncode)
with open(biom_output) as fh:
biom_table = biom.Table.from_tsv(fh, None, None, None)
group_output_df = pd.read_csv(group_output, sep='\t')
return biom_table, group_output_df
