def summarize_analysis(
*args,
**kwargs) -> Tuple[Optional[Sample], Counter[Id], Counter[Id], Scores]:
"""
Summarize for a cross-analysis (to be usually called in parallel!).
"""
# Recover input and parameters
analysis: str = args[0]
ontology: Ontology = kwargs['ontology']
# TODO: Delete the following comment lines in a future release
# including = ontology.including # See comment below for the reason
# excluding = ontology.excluding # in/excluding are not used anymore
counts: Dict[Sample, Counter[Id]] = kwargs['counts']
scores: Dict[Sample, Dict[Id, Score]] = kwargs['scores']
samples: List[Sample] = kwargs['samples']
output: io.StringIO = io.StringIO(newline='')
# Declare/define variables
summary_counts: Counter[Id] = col.Counter()
summary_acc: Counter[Id] = col.Counter()
summary_score: Scores = Scores({})
summary: Optional[Sample] = None
output.write(gray('Summary for ') + analysis + gray('... '))
target_samples: List[Sample] = [
smpl for smpl in samples if smpl.startswith(analysis)
]
assert len(target_samples) >= 1, \
red('ERROR! ') + analysis + gray(' has no samples to summarize!')
for smpl in target_samples:
summary_counts += counts[smpl]
summary_score.update(scores[smpl])
tree = TaxTree()
tree.grow(ontology=ontology, counts=summary_counts, scores=summary_score)
tree.subtract()
tree.shape()
summary_counts.clear()
summary_score.clear()
# Avoid including/excluding here as get_taxa is not as 'clever' as allin1
# and taxa are already included/excluded in the derived samples
tree.get_taxa(counts=summary_counts,
accs=summary_acc,
scores=summary_score)
summary_counts = +summary_counts # remove counts <= 0
if summary_counts: # Avoid returning empty sample (summary would be None)
summary = Sample(f'{analysis}_{STR_SUMMARY}')
output.write(
gray('(') + cyan(f'{len(target_samples)}') + gray(' samples)') +
green(' OK!\n'))
else:
output.write(yellow(' VOID\n'))
# Print output and return
print(output.getvalue(), end='')
sys.stdout.flush()
return summary, summary_counts, summary_acc, summary_score
def summarize_analysis(
*args, **kwargs) -> Tuple[Sample, Counter[Id], Counter[Id], Scores]:
"""
Summarize for a cross-analysis (to be usually called in parallel!).
"""
# Recover input and parameters
analysis: str = args[0]
ontology: Ontology = kwargs['ontology']
including = ontology.including
excluding = ontology.excluding
counts: Dict[Sample, Counter[Id]] = kwargs['counts']
scores: Dict[Sample, Dict[Id, Score]] = kwargs['scores']
samples: List[Sample] = kwargs['samples']
output: io.StringIO = io.StringIO(newline='')
# Declare/define variables
summary_counts: Counter[Id] = Counter()
summary_acc: Counter[Id] = Counter()
summary_score: Scores = Scores({})
summary: Sample = None
output.write(gray('Summary for ') + analysis + gray('... '))
target_samples: List[Sample] = [
smpl for smpl in samples if smpl.startswith(analysis)
]
assert len(target_samples) >= 1, \
red('ERROR! ') + analysis + gray(' has no samples to summarize!')
for smpl in target_samples:
summary_counts += counts[smpl]
summary_score.update(scores[smpl])
tree = TaxTree()
tree.grow(ontology=ontology, counts=summary_counts, scores=summary_score)
tree.subtract()
tree.shape()
summary_counts.clear()
summary_score.clear()
tree.get_taxa(counts=summary_counts,
accs=summary_acc,
scores=summary_score,
include=including,
exclude=excluding)
summary_counts = +summary_counts # remove counts <= 0
if summary_counts: # Avoid returning empty sample (summary would be None)
summary = Sample(f'{analysis}_{STR_SUMMARY}')
output.write(
gray('(') + cyan(f'{len(target_samples)}') + gray(' samples)') +
green(' OK!\n'))
else:
output.write(yellow(' VOID\n'))
# Print output and return
print(output.getvalue(), end='')
sys.stdout.flush()
return summary, summary_counts, summary_acc, summary_score
def process_report(
*args, **kwargs
) -> Tuple[Sample, TaxTree, SampleDataByTaxId, SampleStats, Err]:
"""
Process Centrifuge/Kraken report files (to be usually called in parallel!).
"""
# TODO: Full review to report support
# Recover input and parameters
filerep: Filename = args[0]
taxonomy: Taxonomy = kwargs['taxonomy']
mintaxa: int = kwargs['mintaxa']
collapse: bool = taxonomy.collapse
including: Set[TaxId] = taxonomy.including
excluding: Set[TaxId] = taxonomy.excluding
debug: bool = kwargs['debug']
output: io.StringIO = io.StringIO(newline='')
def vwrite(*args):
"""Print only if verbose/debug mode is enabled"""
if kwargs['debug']:
output.write(' '.join(str(item) for item in args))
sample: Sample = Sample(filerep)
# Read Centrifuge/Kraken report file to get abundances
log: str
abundances: Counter[TaxId]
log, abundances, _ = read_report(filerep)
output.write(log)
# Remove root counts, in case
if kwargs['root']:
vwrite(gray('Removing'), abundances[ROOT], gray('"ROOT" reads... '))
abundances[ROOT] = 0
vwrite(green('OK!'), '\n')
# Build taxonomy tree
output.write(' \033[90mBuilding taxonomy tree...\033[0m')
tree = TaxTree()
tree.grow(taxonomy=taxonomy,
counts=abundances) # Grow tax tree from root node
output.write('\033[92m OK! \033[0m\n')
# Prune the tree
output.write(' \033[90mPruning taxonomy tree...\033[0m')
tree.prune(mintaxa, None, collapse, debug)
tree.shape()
output.write('\033[92m OK! \033[0m\n')
# Get the taxa with their abundances and taxonomical levels
output.write(' \033[90mFiltering taxa...\033[0m')
new_abund: Counter[TaxId] = col.Counter()
new_accs: Counter[TaxId] = col.Counter()
ranks: Ranks = Ranks({})
tree.get_taxa(abundance=new_abund,
accs=new_accs,
ranks=ranks,
mindepth=0,
maxdepth=0,
include=including,
exclude=excluding)
new_abund = +new_abund # remove zero and negative counts
if including or excluding: # Recalculate accumulated counts
new_tree = TaxTree()
new_tree.grow(taxonomy, new_abund) # Grow tree with new abund
new_tree.shape()
new_abund = col.Counter() # Reset abundances
new_accs = col.Counter() # Reset accumulated
new_tree.get_taxa(new_abund, new_accs) # Get new accumulated counts
out: SampleDataByTaxId = SampleDataByTaxId()
out.set(counts=new_abund, ranks=ranks, accs=new_accs)
output.write('\033[92m OK! \033[0m\n')
print(output.getvalue())
sys.stdout.flush()
return sample, tree, out, SampleStats(), Err.NO_ERROR