def _summarize_combined(samples, vkey):
"""Prepare summarized CSV and plot files for samples to combine together.
Helps handle cases where we want to summarize over multiple samples.
"""
validate_dir = utils.safe_makedir(os.path.join(samples[0]["dirs"]["work"], vkey))
combined, _ = _group_validate_samples(samples, vkey, [["metadata", "validate_combine"]])
for vname, vitems in combined.items():
if vname:
cur_combined = collections.defaultdict(int)
for data in sorted(vitems, key=lambda x: x.get("lane", dd.get_sample_name(x))):
validations = [variant.get(vkey) for variant in data.get("variants", [])]
validations = [v for v in validations if v]
if len(validations) == 0 and vkey in data:
validations = [data.get(vkey)]
for validate in validations:
with open(validate["summary"]) as in_handle:
reader = csv.reader(in_handle)
next(reader) # header
for _, caller, vtype, metric, value in reader:
cur_combined[(caller, vtype, metric)] += int(value)
out_csv = os.path.join(validate_dir, "grading-summary-%s.csv" % vname)
with open(out_csv, "w") as out_handle:
writer = csv.writer(out_handle)
header = ["sample", "caller", "vtype", "metric", "value"]
writer.writerow(header)
for (caller, variant_type, category), val in cur_combined.items():
writer.writerow(["combined-%s" % vname, caller, variant_type, category, val])
plots = validateplot.classifyplot_from_valfile(out_csv)
def evaluate(data):
"""Provide evaluations for multiple callers split by structural variant type.
"""
work_dir = utils.safe_makedir(
os.path.join(data["dirs"]["work"], "structural",
dd.get_sample_name(data), "validate"))
truth_sets = tz.get_in(["config", "algorithm", "svvalidate"], data)
if truth_sets and data.get("sv"):
if isinstance(truth_sets, dict):
val_summary, df_csv = _evaluate_multi(data["sv"], truth_sets,
work_dir, data)
summary_plots = _plot_evaluation(df_csv)
data["sv-validate"] = {
"csv": val_summary,
"plot": summary_plots,
"df": df_csv
}
else:
assert isinstance(
truth_sets,
basestring) and utils.file_exists(truth_sets), truth_sets
val_summary = _evaluate_vcf(data["sv"], truth_sets, work_dir, data)
title = "%s structural variants" % dd.get_sample_name(data)
summary_plots = validateplot.classifyplot_from_valfile(
val_summary, outtype="png", title=title)
data["sv-validate"] = {
"csv": val_summary,
"plot": summary_plots[0]
}
return data
def evaluate(data):
"""Provide evaluations for multiple callers split by structural variant type.
"""
work_dir = utils.safe_makedir(os.path.join(data["dirs"]["work"], "structural",
dd.get_sample_name(data), "validate"))
truth_sets = tz.get_in(["config", "algorithm", "svvalidate"], data)
if truth_sets and data.get("sv"):
if isinstance(truth_sets, dict):
val_summary, df_csv = _evaluate_multi(data["sv"], truth_sets, work_dir, data)
summary_plots = _plot_evaluation(df_csv)
data["sv-validate"] = {"csv": val_summary, "plot": summary_plots, "df": df_csv}
else:
assert isinstance(truth_sets, basestring) and utils.file_exists(truth_sets), truth_sets
val_summary = _evaluate_vcf(data["sv"], truth_sets, work_dir, data)
title = "%s structural variants" % dd.get_sample_name(data)
summary_plots = validateplot.classifyplot_from_valfile(val_summary, outtype="png", title=title)
data["sv-validate"] = {"csv": val_summary, "plot": summary_plots[0]}
return data
import sys from bcbio.variation import validateplot title = "smCounter2 UMI: VarDict 1.5.6; octopus 0.5.1b" validateplot.classifyplot_from_valfile(sys.argv[1], outtype="png", title=title)
import sys from bcbio.variation import validateplot title="ICR142: GATK HaplotypeCaller, FreeBayes, VarDict, Platypus" validateplot.classifyplot_from_valfile(sys.argv[1], outtype="png", title=title)
