def test_accuracy_as_phred_qv(self):
qv = accuracy_as_phred_qv(0.999)
self.assertEquals(int(round(qv)), 30)
qv = accuracy_as_phred_qv(1.0, max_qv=60)
self.assertEquals(int(round(qv)), 60)
qv = accuracy_as_phred_qv([0.95, 1.0, 0.99999])
qvs = [int(round(x)) for x in qv]
self.assertEqual(qvs, [13, 70, 50])
def test_accuracy_as_phred_qv(self):
qv = accuracy_as_phred_qv(0.999)
self.assertEquals(int(round(qv)), 30)
qv = accuracy_as_phred_qv(1.0, max_qv=60)
self.assertEquals(int(round(qv)), 60)
qv = accuracy_as_phred_qv([0.95, 1.0, 0.99999])
qvs = [ int(round(x)) for x in qv ]
self.assertEqual(qvs, [13, 70, 50])
def _movie_results_to_attributes(movie_results):
"""Create the necessary attributes for the CCS report"""
rs = [m.read_lengths for m in movie_results]
read_lengths = np.concatenate(rs)
ac = [m.accuracies for m in movie_results]
accuracies = np.concatenate(ac)
npass = [m.num_passes for m in movie_results]
num_passes = np.concatenate(npass)
m_readlength = int(read_lengths.mean()) if read_lengths.size > 0 else 0.0
m_accuracy = np.round(
accuracies.mean(), decimals=4) if accuracies.size > 0 else 0.0
m_npasses = np.round(num_passes.mean()) if num_passes.size > 0 else 0.0
m_qv = int(round(accuracy_as_phred_qv(float(m_accuracy))))
n_reads_at = Attribute(
Constants.A_NREADS, read_lengths.shape[0], name=Constants.ATTR_LABELS[
Constants.A_NREADS])
t_bases_at = Attribute(
Constants.A_TOTAL_BASES, read_lengths.sum(), name=Constants.ATTR_LABELS[Constants.A_TOTAL_BASES])
m_readlength_at = Attribute(
Constants.A_MEAN_READLENGTH, m_readlength, name=Constants.ATTR_LABELS[Constants.A_MEAN_READLENGTH])
m_accuracy_at = Attribute(
Constants.A_MEAN_ACCURACY, m_accuracy, name=Constants.ATTR_LABELS[Constants.A_MEAN_ACCURACY])
m_qv = Attribute(
Constants.A_MEAN_QV, m_qv, name=Constants.ATTR_LABELS[Constants.A_MEAN_QV])
m_npasses_at = Attribute(
Constants.A_MEAN_NPASSES, m_npasses, name=Constants.ATTR_LABELS[Constants.A_MEAN_NPASSES])
attributes = [n_reads_at, t_bases_at,
m_readlength_at, m_accuracy_at, m_qv, m_npasses_at]
return attributes
def _movie_results_to_table(movie_results):
"""Group movie results by movie name and build a report table.
Table has movie name, # of CCS bases, Total CCS bases,
mean CCS readlength and mean CCS accuracy.
"""
columns = [Column(Constants.C_MOVIE_NAME, header="Movie"),
Column(Constants.A_NREADS, header="Consensus reads"),
Column(Constants.A_TOTAL_BASES,
header="Number of consensus bases"),
Column(Constants.A_MEAN_READLENGTH,
header="Mean Consensus Read Length"),
Column(Constants.A_MEAN_ACCURACY,
header="Mean Consensus Predicted Accuracy"),
Column(Constants.A_MEAN_QV,
header="Mean Consensus Predicted QV"),
Column(Constants.A_MEAN_NPASSES, header="Mean Number of Passes")]
table = Table(Constants.T_ID, title="Consensus reads", columns=columns)
movie_names = {m.movie_name for m in movie_results}
for movie_name in movie_names:
rs = [
m.read_lengths for m in movie_results if m.movie_name == movie_name]
read_lengths = np.concatenate(rs)
ac = [
m.accuracies for m in movie_results if m.movie_name == movie_name]
accuracies = np.concatenate(ac)
npass = [
m.num_passes for m in movie_results if m.movie_name == movie_name]
num_passes = np.concatenate(npass)
m_readlength = int(
read_lengths.mean()) if read_lengths.size > 0 else 0.0
m_accuracy = np.round(
accuracies.mean(), decimals=4) if accuracies.size > 0 else 0.0
m_npasses = np.round(
num_passes.mean(), decimals=3) if num_passes.size > 0 else 0.0
m_qv = int(round(accuracy_as_phred_qv(float(accuracies.mean()))))
table.add_data_by_column_id(Constants.C_MOVIE_NAME, movie_name)
table.add_data_by_column_id(Constants.A_NREADS, read_lengths.shape[0])
table.add_data_by_column_id(
Constants.A_TOTAL_BASES, read_lengths.sum())
table.add_data_by_column_id(Constants.A_MEAN_READLENGTH, m_readlength)
table.add_data_by_column_id(Constants.A_MEAN_ACCURACY, m_accuracy)
table.add_data_by_column_id(Constants.A_MEAN_QV, m_qv)
table.add_data_by_column_id(Constants.A_MEAN_NPASSES, m_npasses)
return table
def scatter_plot_accuracy_vs_numpasses(data,
axis_labels=("Number of passes", "Predicted accuracy (Phred QV)"),
nbins=None, barcolor=None):
"""
"""
npasses, accuracy = data
qvs = accuracy_as_phred_qv(accuracy)
fig, ax = get_fig_axes_lpr()
data = [Line(xData=npasses,
yData=qvs,
style='o')]
apply_line_data(
ax=ax,
line_models=data,
axis_labels=axis_labels,
only_whole_ticks=False)
return fig, ax
def scatter_plot_accuracy_vs_numpasses(
data,
axis_labels=(
get_plot_xlabel(spec, Constants.PG_SCATTER, Constants.P_SCATTER),
get_plot_ylabel(spec, Constants.PG_SCATTER, Constants.P_SCATTER),
),
nbins=None,
barcolor=None,
):
"""
"""
npasses, accuracy = data
qvs = accuracy_as_phred_qv(accuracy)
fig, ax = get_fig_axes_lpr()
data = [Line(xData=npasses, yData=qvs, style="o")]
apply_line_data(ax=ax, line_models=data, axis_labels=axis_labels, only_whole_ticks=False)
return fig, ax
def scatter_plot_accuracy_vs_numpasses(
data,
axis_labels=(get_plot_xlabel(spec, Constants.PG_SCATTER,
Constants.P_SCATTER),
get_plot_ylabel(spec, Constants.PG_SCATTER,
Constants.P_SCATTER)),
nbins=None,
barcolor=None):
"""
"""
npasses, accuracy = data
qvs = accuracy_as_phred_qv(accuracy)
fig, ax = get_fig_axes_lpr()
data = [Line(xData=npasses, yData=qvs, style='o')]
apply_line_data(ax=ax,
line_models=data,
axis_labels=axis_labels,
only_whole_ticks=False)
return fig, ax
def scatter_plot_accuracy_vs_numpasses(data,
axis_labels=(
meta_rpt.get_meta_plotgroup(Constants.PG_SCATTER).get_meta_plot(Constants.P_SCATTER).xlabel,
meta_rpt.get_meta_plotgroup(Constants.PG_SCATTER).get_meta_plot(Constants.P_SCATTER).ylabel),
nbins=None, barcolor=None):
"""
"""
npasses, accuracy = data
qvs = accuracy_as_phred_qv(accuracy)
fig, ax = get_fig_axes_lpr()
data = [Line(xData=npasses,
yData=qvs,
style='o')]
apply_line_data(
ax=ax,
line_models=data,
axis_labels=axis_labels,
only_whole_ticks=False)
return fig, ax