def plot_ranks(self, filename=None, savefig=False):
# ranks
# the *score* columns contains the scaled IDR value, min(int(log2(-125IDR), 1000).
# e.g. peaks with an IDR of 0 have a score of 1000, idr 0.05 have a score of
# int(-125log2(0.05)) = 540, and idr 1.0 has a score of 0.
df1 = self.df.query('score>540')
df2 = self.df.query('score<=540')
pylab.clf()
pylab.plot(df1.rep1_rank,
df1.rep2_rank,
'ko',
alpha=0.5,
label='<0.05 IDR')
pylab.plot(df2.rep1_rank,
df2.rep2_rank,
'ro',
alpha=0.5,
label='>=0.05 IDR')
pylab.xlabel("Peak rank - replicate 1")
pylab.ylabel("Peak rank - replicate 2")
N = len(self.df)
pylab.plot([0, N], [0, N], color='blue', alpha=0.5, ls='--')
#pylab.xlim([0,1.05])
#pylab.ylim([0,1.05])
pylab.legend(loc='lower right')
if savefig:
pylab.savefig(filename)
def plot_unknown_barcodes(self, N=20):
ub = self.data['UnknownBarcodes']
df = pd.DataFrame({x['Lane']: x['Barcodes'] for x in ub})
if "unknown" in df.index and len(df) == 1:
df.loc['known'] = [0 for i in df.columns]
# if data is made of undetermined only, the dataframe is just made of
# N lanes with one entry : unknown
S = df.sum(axis=1).sort_values(ascending=False).index[0:N]
data = df.loc[S][::-1]
#print(data)
data.columns = ["Lane {}".format(x) for x in data.columns]
from matplotlib import rcParams
rcParams['axes.axisbelow'] = True
pylab.figure(figsize=(10, 8))
ax = pylab.gca()
data.plot(kind="barh", width=1, ec="k", ax=ax)
rcParams['axes.axisbelow'] = False
pylab.xlabel("Number of reads", fontsize=12)
pylab.ylabel("")
pylab.grid(True)
pylab.legend(
["Lane {}".format(x) for x in range(1,
len(df.columns) + 1)],
loc="lower right")
try:
pylab.tight_layout()
except Exception as err:
print(err)
return data
def hist_ORF_CDS_linearscale(self,
alpha=0.5,
bins=40,
xlabel="Length",
ylabel="#"):
if self._ORF_pos is None:
self._find_ORF_CDS()
n_ORF = self._ORF_pos["len_ORF"].dropna().shape[0]
n_CDS = self._ORF_pos["len_CDS"].dropna().shape[0]
# plot for all ORF and CDS
pylab.hist(self._ORF_pos["len_ORF"].dropna(),
alpha=alpha,
label="ORF, N = " + str(n_ORF),
bins=bins)
pylab.hist(self._ORF_pos["len_CDS"].dropna(),
alpha=alpha,
label="CDS, N = " + str(n_CDS),
bins=bins)
pylab.xlabel(xlabel)
pylab.ylabel(ylabel)
pylab.legend()
pylab.title("Length of ORF and CDS (after filter %s > %d)" \
%(self._type_filter, self._threshold))
def plot_dispersion(self):
pylab.plot(
self.dds_stats.baseMean,
self.dds_stats.dispGeneEst,
"ok",
label="Estimate",
ms=1,
)
pylab.plot(
self.dds_stats.baseMean,
self.dds_stats.dispersion,
"ob",
label="final",
ms=1,
)
pylab.plot(self.dds_stats.baseMean,
self.dds_stats.dispFit,
"or",
label="Fit",
ms=1)
pylab.legend()
ax = pylab.gca()
ax.set(yscale="log")
ax.set(xscale="log")
self._format_plot(
title="Dispersion estimation",
xlabel="Mean of normalized counts",
ylabel="Dispersion",
)
def plot_alignment(self, bamfile, motif, window=200,
global_th=10,title=None,legend=True, legend_fontsize=11,
valid_rnames=[],
valid_flags=[]):
"""
plot alignments that match the motif.
"""
bam = BAM(bamfile)
print("Found {} hits".format(len(bam)))
pylab.clf()
count = 0
for aln in bam:
if valid_rnames and aln.rname not in valid_rnames:
continue
if valid_flags and aln.flag not in valid_flags:
continue
seq = aln.query_sequence
if seq:
count += 1
X1 = [seq[i:i+window].count(motif) for i in range(len(seq))]
pylab.plot(range(aln.reference_start,
aln.reference_start+len(seq)),X1, label=aln.query_name)
print("Showing {} entries after filtering".format(count))
max_theo = int(1.2*window / len(motif))
pylab.ylim([0, max_theo])
if legend and count<15:
pylab.legend(fontsize=legend_fontsize)
if title:
pylab.title(title, fontsize=16)
def plot_specific_alignment(self, bamfile, query_name, motif,clf=True,
show_figure=True, authorized_flags=[0,16],
windows=[10, 50, 100, 150,200, 250,500, 1000], local_threshold=5):
found = None
bam = BAM(bamfile)
for aln in bam:
if aln.query_name == query_name and aln.flag in authorized_flags:
found = aln
break # we may have several entries. let us pick up the first
sizes = []
if found:
# Detection
seq = found.query_sequence
if clf:pylab.clf()
for window in windows:
X = [seq[i:i+window].count(motif) for i in range(len(seq))]
if show_figure:
pylab.plot(X, label=window)
score = sum([x>local_threshold for x in X])
sizes.append(score-window)
if show_figure:
pylab.legend()
pylab.ylabel("# {} in a given sliding window".format(motif))
pylab.title(query_name)
else:
print("{} Not found in {} file".format(query_name, bamfile))
return sizes
def plot_stacked_hist(self,
output_filename=None,
dpi=200,
kind="barh",
fontsize=10,
edgecolor="k",
lw=1,
width=1,
ytick_fontsize=10):
df = self.get_df()
df.T.plot(kind=kind,
stacked=True,
edgecolor=edgecolor,
lw=lw,
width=width)
ax = pylab.gca()
positions = pylab.yticks()
#ax.set_yticklabel(positions, labels, fontsize=ytick_fontsize)
pylab.xlabel("Percentage (%)", fontsize=fontsize)
pylab.ylabel("Sample index/name", fontsize=fontsize)
pylab.yticks(fontsize=ytick_fontsize)
pylab.legend(title="kingdom")
pylab.xlim([0, 100])
if output_filename:
pylab.savefig(output_filename, dpi=dpi)
def scatter_plot(self, filename=None, hold=False):
"""Scatter plot of the score versus length of each ortholog
.. plot::
:include-source:
from sequana import BUSCO, sequana_data
b = BUSCO(sequana_data("test_busco_full_table.tsv"))
b.scatter_plot()
"""
if hold is False:
pylab.clf()
colors = ["green", "orange", "red", "blue"]
markers = ['o', 's', 'x', 'o']
for i, this in enumerate(["Complete", "Fragmented", "Missing", "Duplicated"]):
mask = self.df.Status == "Complete"
if sum(mask)>0:
self.df[mask].plot(x="Length", y="Score", kind="scatter",
color=colors[i],
marker=markers[i], label="Complete")
pylab.legend()
pylab.grid()
if filename:
pylab.savefig(filename)
def hist_average_quality(self, fontsize=16, bins=None):
"""
bins is from 0 to 94
"""
hq_qv = [pylab.mean([ord(X)-33 for X in read['quality'].decode()])
for read in self.hq_sequence]
lq_qv = [pylab.mean([ord(X) -33 for X in read['quality'].decode()])
for read in self.lq_sequence]
if bins is None:
bins = range(0,94)
Y1, X = np.histogram(hq_qv, bins=bins)
Y2, X = np.histogram(lq_qv, bins=bins)
pylab.bar(X[1:], Y1, width=1, label="HQ")
pylab.bar(X[1:], Y2, bottom=Y1, width=1, label="LQ")
pylab.xlim([0.5, 93.5])
pylab.xlabel("Isoform average QV")
pylab.ylabel("# Isoform")
pylab.legend(fontsize=fontsize)
ax = pylab.twinx()
N = np.sum(Y1+Y2)
ax.plot(X, [N] + list(N-np.cumsum(Y1+Y2)), "k")
def plot_polymerase_per_barcode(self, fontsize=12, unbarcoded=True):
"""Number Of Polymerase Reads Per Barcode"""
PR = self.df_barcoded["Polymerase Reads"].sum()
data = self.df_barcoded['Polymerase Reads'].sort_values(
ascending=False).values
pylab.plot([int(x) for x in range(1,
len(data) + 1)],
data,
label="barcodes")
pylab.axhline(data.mean(), color="r", label="average")
try:
if unbarcoded is True:
unbar = self.df_not_barcoded['Polymerase Reads'].iloc[0]
pylab.axhline(unbar, color="k", ls="--", label="not barcoded")
except:
pass
pylab.xlabel("Barcode Rank Order", fontsize=fontsize)
pylab.ylabel("Counts of Reads", fontsize=fontsize)
pylab.title("Total Polymerase count: {}".format(PR))
pylab.legend()
pylab.ylim(ymin=0)
try:
pylab.tight_layout()
except:
pass
def barplot_count_ORF_CDS_by_frame(self,
alpha=0.5,
bins=40,
xlabel="Frame",
ylabel="#",
bar_width=0.35):
if self._ORF_pos is None:
self._find_ORF_CDS()
# number of ORF and CDS found by frame
frames = [-3, -2, -1, 1, 2, 3]
nb_res_ORF = []
nb_res_CDS = []
for fr in frames:
nb_res_ORF.append(self._ORF_pos[self._ORF_pos["frame"] == fr]
["len_ORF"].dropna().shape[0])
nb_res_CDS.append(self._ORF_pos[self._ORF_pos["frame"] == fr]
["len_CDS"].dropna().shape[0])
pylab.bar(np.array(frames) - (bar_width / 2),
nb_res_ORF,
bar_width,
alpha=alpha,
label="ORF N = %d" % sum(nb_res_ORF))
pylab.bar(np.array(frames) + (bar_width / 2),
nb_res_CDS,
bar_width,
alpha=alpha,
label="CDS N = %d" % sum(nb_res_CDS))
pylab.xlabel(xlabel)
pylab.ylabel(ylabel)
pylab.legend(loc=1)
pylab.title("Number of ORF and CDS by frame")
def plot_specific_alignment(self,
query_name,
motif,
clf=True,
windows=[10, 50, 100, 200, 500, 1000]):
found = None
bam = BAM(self.bamfile)
for aln in bam:
if aln.query_name == query_name:
found = aln
if found:
# Detection
seq = found.query_sequence
if clf: pylab.clf()
for window in windows:
X = [seq[i:i + window].count(motif) for i in range(len(seq))]
pylab.plot(X, label=window)
score = sum([x > window / 6 for x in X])
print(window, score / 3.)
pylab.legend()
pylab.ylabel("# {} in a given sliding window".format(motif))
pylab.title(query_name)
else:
print("Not found")
def scatter_plot(self, filename=None, hold=False):
"""Scatter plot of the score versus length of each ortholog
.. plot::
:include-source:
from sequana import BUSCO, sequana_data
b = BUSCO(sequana_data("test_busco_full_table.tsv"))
b.scatter_plot()
Missing are not show since there is no information about contig .
"""
if hold is False:
pylab.clf()
colors = ["green", "orange", "red", "blue"]
markers = ['o', 's', 'x', 'o']
for i, this in enumerate(["Complete", "Fragmented", "Duplicated"]):
mask = self.df.Status == this
if sum(mask) > 0:
self.df[mask].plot(x="Length",
y="Score",
kind="scatter",
color=colors[i],
ax=pylab.gca(),
marker=markers[i],
label=this)
pylab.legend()
pylab.grid()
if filename:
pylab.savefig(filename)
def hist_length_repeats(self,
bins=None,
alpha=0.5,
hold=False,
fontsize=12,
grid=True,
label="Repeat length",
xlabel="Repeat length",
ylabel="#"):
"""Plots histogram of the repeat lengths
"""
# check that user has set a threshold
if self._list_len_repeats is None:
self._get_list_len_repeats()
if bins is None:
bins = range(max(0, self.threshold - 1),
max(self._list_len_repeats) + 2)
if hold is False:
pylab.clf()
pylab.hist(self._list_len_repeats, alpha=alpha, label=label, bins=bins)
pylab.legend()
pylab.xlabel(xlabel, fontsize=fontsize)
pylab.ylabel(ylabel, fontsize=fontsize)
if grid is True:
pylab.grid(True)
def hist_average_quality(self, fontsize=16, bins=None):
"""
bins is from 0 to 94
"""
hq_qv = [pylab.mean([ord(X)-33 for X in read['quality'].decode()])
for read in self.hq_sequence]
lq_qv = [pylab.mean([ord(X) -33 for X in read['quality'].decode()])
for read in self.lq_sequence]
if bins is None:
bins = range(0,94)
Y1, X = np.histogram(hq_qv, bins=bins)
Y2, X = np.histogram(lq_qv, bins=bins)
pylab.bar(X[1:], Y1, width=1, label="HQ")
pylab.bar(X[1:], Y2, bottom=Y1, width=1, label="LQ")
pylab.xlim([0.5, 93.5])
pylab.xlabel("Isoform average QV")
pylab.ylabel("# Isoform")
pylab.legend(fontsize=fontsize)
ax = pylab.twinx()
N = np.sum(Y1+Y2)
ax.plot(X, [N] + list(N-np.cumsum(Y1+Y2)), "k")
def barplot_summary(self,
filename=None,
color=["green", "red"],
alpha=0.8):
df = self.get_data_reads()
under = df.query("name=='Undetermined'")
total = df.query("name!='Undetermined'")
total = total.groupby("lane").sum().reset_index()
total["name"] = "Determined"
df = pd.concat([under, total]) #sort=True)
df = df.pivot(index="lane", columns="name", values="count")
df = df[["Determined", "Undetermined"]]
if df.sum().min() > 1e6:
df /= 1e6
df.plot.barh(stacked=True, color=color, alpha=alpha, ec='k')
pylab.xlabel("Number of reads (M)")
else:
df.plot.barh(stacked=True, color=color, alpha=alpha, ec='k')
pylab.xlabel("Number of reads")
pylab.legend()
if filename:
pylab.savefig(filename, dpi=200)
return df
def hist_snr(self, bins=50, alpha=0.5, hold=False, fontsize=12,
grid=True, xlabel="SNR", ylabel="#",title="", clip_upper_SNR=30):
"""Plot histogram of the ACGT SNRs for all reads
:param int bins: binning for the histogram. Note that the range starts
at 0 and ends at clip_upper_SNR
:param float alpha: transparency of the histograms
:param bool hold:
:param int fontsize:
:param bool grid:
:param str xlabel:
:param str ylabel:
:param str title:
.. plot::
:include-source:
from sequana.pacbio import PacbioSubreads
from sequana import sequana_data
b = PacbioSubreads(sequana_data("test_pacbio_subreads.bam"))
b.hist_snr()
"""
if self._df is None:
self._get_df()
# old pacbio format has no SNR stored
if len(self._df['snr_A'].dropna()) == 0:
# nothing to plot
from sequana import sequana_data
pylab.clf()
pylab.imshow(pylab.imread(sequana_data("no_data.jpg")))
pylab.gca().axis('off')
return
if hold is False:
pylab.clf()
maxSNR = 0
for letter in "ACGT":
m = self._df.loc[:,"snr_{}".format(letter)].max()
if m > maxSNR:
maxSNR = m
if maxSNR > clip_upper_SNR:
maxSNR = clip_upper_SNR
bins = pylab.linspace(0, maxSNR, bins)
pylab.hist(self._df.loc[:,'snr_A'].clip_upper(maxSNR), alpha=alpha, label="A", bins=bins)
pylab.hist(self._df.loc[:,'snr_C'].clip_upper(maxSNR), alpha=alpha, label="C", bins=bins)
pylab.hist(self._df.loc[:,'snr_G'].clip_upper(maxSNR), alpha=alpha, label="G", bins=bins)
pylab.hist(self._df.loc[:,'snr_T'].clip_upper(maxSNR), alpha=alpha, label="T", bins=bins)
pylab.legend()
pylab.xlabel(xlabel, fontsize=fontsize)
pylab.ylabel(ylabel, fontsize=fontsize)
pylab.title(title,fontsize=fontsize)
if grid is True:
pylab.grid(True)
def plot(self, chr_name, x1=None, x2=None, Y=20):
df = self.df.query("name == @chr_name")
for _, item in df.iterrows():
if item['type'] == "deletion":
plot([item.start, item.end], [-1, -1], "r-", label="deletion")
else:
plot([item.start, item.end], [Y, Y], "b-", label="duplication")
pylab.legend()
def barplot_per_sample(self, alpha=0.5, width=0.8, filename=None):
df = self.get_data_reads()
# this is ugly but will do the job for now
under = df.query("name=='Undetermined'")
others = df.query("name!='Undetermined'")
under = under.groupby("name").sum().reset_index()
others = others.groupby("name").sum().reset_index()
under = under[["name", "count"]].set_index("name")
others = others[["name", "count"]].set_index("name")
all_data = others.sort_index(ascending=False)
all_data.columns = ["samples"]
# appended at the end
all_data.loc['undetermined'] = 0
# revert back
all_data = all_data.loc[::-1]
# just for legend
under.columns = ['undetermined']
if all_data.sum().min() > 1e6:
all_data /= 1e6
under /= 1e6
M = True
else:
M = False
all_data.plot(kind="barh", alpha=alpha, zorder=1, width=width, ec='k')
under.plot(kind="barh",
alpha=alpha,
color="red",
ax=pylab.gca(),
zorder=1,
width=width,
ec='k')
pylab.ylim([-0.5, len(all_data) + 0.5])
if len(all_data) < 100:
pylab.yticks(range(len(all_data)), all_data.index)
pylab.legend()
pylab.grid(True, zorder=-1)
if M:
pylab.xlabel("Number of reads (M)")
else:
pylab.xlabel("Number of reads")
try:
pylab.tight_layout()
except:
pass
if filename:
pylab.savefig(filename, dpi=200)
def plot_volcano_differences(self, mode="all"):
cond1, cond2 = "cond1", "cond2"
labels = [cond1, cond2]
A = self.r1.df.loc[self.r1.gene_lists[mode]]
B = self.r2.df.loc[self.r2.gene_lists[mode]]
AB = set(A.index).intersection(set(B.index))
Aonly = A.loc[set(A.index).difference(set(B.index))]
Bonly = B.loc[set(B.index).difference(set(A.index))]
Acommon = A.loc[AB]
Bcommon = B.loc[AB]
pylab.clf()
pylab.plot(Acommon.log2FoldChange, -np.log10(Acommon.padj), marker="o",
alpha=0.5, color="r", lw=0, label="Common in experiment 1", pickradius=4,
picker=True)
pylab.plot(Bcommon.log2FoldChange, -np.log10(Bcommon.padj), marker="o",
alpha=0.5, color="orange", lw=0, label="Common in experiment 2", pickradius=4,
picker=True)
for x in AB:
a_l = A.loc[x].log2FoldChange
a_p = -np.log10(A.loc[x].padj)
b_l = B.loc[x].log2FoldChange
b_p = -np.log10(B.loc[x].padj)
pylab.plot([a_l, b_l], [a_p, b_p], 'k', alpha=0.5)
pylab.plot(Bonly.log2FoldChange, -np.log10(Bonly.padj), marker="*",
alpha=0.5, color="blue", lw=0, label="In experiment 2 only", pickradius=4,
picker=True)
pylab.plot(Aonly.log2FoldChange, -np.log10(Aonly.padj), marker="*",
alpha=0.5, color="cyan", lw=0, label="In experiment 1 only", pickradius=4,
picker=True)
for name, x in Bonly.iterrows():
x1 = x.log2FoldChange
y1 = -np.log10(x.padj)
x2 = self.r1.df.loc[name].log2FoldChange
y2 = -np.log10(self.r1.df.loc[name].padj)
pylab.plot( [x1,x2], [y1,y2], ls="--", color='r')
for name, x in Aonly.iterrows():
x1 = x.log2FoldChange
y1 = -np.log10(x.padj)
x2 = self.r2.df.loc[name].log2FoldChange
y2 = -np.log10(self.r2.df.loc[name].padj)
pylab.plot( [x1,x2], [y1,y2], ls="-", color='r')
pylab.axhline(1.33, alpha=0.5, ls="--", color="r")
pylab.xlabel("log2 fold Change")
pylab.ylabel("log10 adjusted p-values")
pylab.legend()
pylab.grid(True)
return Aonly, Bonly, Acommon, Bcommon
def hist_snr(self,
bins=50,
alpha=0.5,
hold=False,
fontsize=12,
grid=True,
xlabel="SNR",
ylabel="#",
title=""):
"""Plot histogram of the ACGT SNRs for all reads
:param int bins: binning for the histogram
:param float alpha: transparency of the histograms
:param bool hold:
:param int fontsize:
:param bool grid:
:param str xlabel:
:param str ylabel:
:param str title:
.. plot::
:include-source:
from sequana.pacbio import BAMPacbio
from sequana import sequana_data
b = BAMPacbio(sequana_data("test_pacbio_subreads.bam"))
b.hist_snr()
"""
if self._df is None:
self._get_df()
# old pacbio format has no SNR stored
if len(self._df['snr_A'].dropna()) == 0:
# nothing to plot
from sequana import sequana_data
pylab.clf()
pylab.imshow(pylab.imread(sequana_data("no_data.jpg")))
pylab.gca().axis('off')
return
if hold is False:
pylab.clf()
pylab.hist(self._df.loc[:, 'snr_A'], alpha=alpha, label="A", bins=bins)
pylab.hist(self._df.loc[:, 'snr_C'], alpha=alpha, label="C", bins=bins)
pylab.hist(self._df.loc[:, 'snr_G'], alpha=alpha, label="G", bins=bins)
pylab.hist(self._df.loc[:, 'snr_T'], alpha=alpha, label="T", bins=bins)
pylab.legend()
pylab.xlabel(xlabel, fontsize=fontsize)
pylab.ylabel(ylabel, fontsize=fontsize)
pylab.title(title, fontsize=fontsize)
if grid is True:
pylab.grid(True)
def hist_isoform_length_mapped_vs_unmapped(self, bins=None):
df = self.df
if bins is None:
bins = range(0, len(df.reference_length.max()), 100)
mapped = df[df.reference_name != -1]
unmapped = df[df.reference_name == -1]
pylab.hist(mapped.reference_length, bins=bins, alpha=0.5,
label="mapped {}".format(len(mapped)), density=False)
pylab.hist(unmapped.reference, bins=bins, alpha=0.5,
label="unmapped {}".format(len(unmapped)), density=False)
pylab.xlabel("Isoform length")
pylab.legend()
def hist_isoform_length_mapped_vs_unmapped(self, bins=None):
df = self.df
if bins is None:
bins = range(0, df.read_length.max(), 100)
mapped = df[df.reference_name != -1]
unmapped = df[df.reference_name == -1]
pylab.hist(mapped.read_length, bins=bins, alpha=0.5,
label="mapped {}".format(len(mapped)), normed=True)
pylab.hist(unmapped.read_length, bins=bins, alpha=0.5,
label="unmapped {}".format(len(unmapped)), normed=True)
pylab.xlabel("Isoform length")
pylab.legend()
def run(self,
bins=50,
xmin=0,
xmax=30000,
step=1000,
burn=1000,
alpha=1,
output_filename=None):
# compute histogram of the input reads once for all to be used
# in the target_distribution method
self.bins = bins
self.Y, self.X = np.histogram(self.bam.df.read_length,
bins=bins,
normed=True)
lengths = self.bam_simul.df.read_length.values
self.tokeep = []
vec = []
x = self.bam.df.read_length.mean()
for i in range(self.bam_simul.df.shape[0]):
can = lengths[i]
aprob = min([
alpha,
self.target_distribution(can) / self.target_distribution(x)
])
#acceptance probability
u = pylab.uniform(0, 1)
if u < aprob:
x = can
vec.append(x)
self.tokeep.append(True)
else:
self.tokeep.append(False)
#plotting the results:
#theoretical curve
x = pylab.arange(xmin, xmax, step)
y = self.target_distribution(x)
pylab.subplot(211)
pylab.title('Metropolis-Hastings')
pylab.plot(vec)
pylab.subplot(212)
pylab.hist(vec[burn:], bins=bins, normed=1)
pylab.plot(x, y, 'r-')
pylab.ylabel('Frequency')
pylab.xlabel('x')
pylab.legend(('PDF', 'Samples'))
if output_filename is not None:
self.bam_simul.filter_bool(output_filename, self.tokeep)
def plot_indel_dist(self, fontsize=16):
"""Plot indel count (+ ratio)
:Return: list of insertions, deletions and ratio insertion/deletion for
different length starting at 1
.. plot::
:include-source:
from sequana import sequana_data, BAM
b = BAM(sequana_data("measles.fa.sorted.bam"))
b.plot_indel_dist()
What you see on this figure is the presence of 10 insertions of length
1, 1 insertion of length 2 and 3 deletions of length 1
# Note that in samtools, several insertions or deletions in a single
alignment are ignored and only the first one seems to be reported. For
instance 10M1I10M1I stored only 1 insertion in its report; Same comment
for deletions.
.. todo:: speed up and handle long reads cases more effitiently by
storing INDELS as histograms rather than lists
"""
try:
self.insertions
except:
self._set_indels()
if len(self.insertions) ==0 or len(self.deletions) == 0:
raise ValueError("No deletions or insertions found")
N = max(max(Counter(self.deletions)), max(Counter(self.insertions))) + 1
D = [self.deletions.count(i) for i in range(N)]
I = [self.insertions.count(i) for i in range(N)]
R = [i/d if d!=0 else 0 for i,d in zip(I, D)]
fig, ax = pylab.subplots()
ax.plot(range(N), I, marker="x", label="Insertions")
ax.plot(range(N), D, marker="x", label="Deletions")
ax.plot(range(N), R, "--r", label="Ratio insertions/deletions")
ax.set_yscale("symlog")
pylab.ylim([1, pylab.ylim()[1]])
pylab.legend()
pylab.grid()
from matplotlib.ticker import MaxNLocator
ax.xaxis.set_major_locator(MaxNLocator(integer=True))
pylab.xlabel("Indel length", fontsize=fontsize)
pylab.ylabel("Indel count", fontsize=fontsize)
return I, D, R
def bar_plot_contigs_length(self):
# show length of N contigs as compare to length of the reference
fref = FastA(self.reference)
Nref = len(fref.sequences)
N = len(self.fasta)
pylab.clf()
pylab.bar(range(0, N, int(pylab.ceil(N / Nref))),
sorted(fref.lengths),
width=Nref / 1.1,
label="Plasmodium chromosomes")
pylab.bar(range(0, N),
sorted(self.fasta.lengths),
width=1,
label="canu {} contigs".format(N))
pylab.legend()
def hist_ORF_CDS_linearscale(self, alpha=0.5, bins=40, xlabel="Length", ylabel="#"):
if self._ORF_pos is None:
self._find_ORF_CDS()
n_ORF = self._ORF_pos["len_ORF"].dropna().shape[0]
n_CDS = self._ORF_pos["len_CDS"].dropna().shape[0]
# plot for all ORF and CDS
pylab.hist(self._ORF_pos["len_ORF"].dropna(),alpha=alpha, label="ORF, N = " + str(n_ORF),bins=bins)
pylab.hist(self._ORF_pos["len_CDS"].dropna(),alpha=alpha, label="CDS, N = " + str(n_CDS),bins=bins)
pylab.xlabel(xlabel)
pylab.ylabel(ylabel)
pylab.legend()
pylab.title("Length of ORF and CDS (after filter %s > %d)" \
%(self._type_filter, self._threshold))
def hist_snr(self, bins=50, alpha=0.5, hold=False, fontsize=12,
grid=True,xlabel="SNR",ylabel="#"):
"""Plot histogram of the ACGT SNRs for all reads"""
if self._df is None:
self._get_df()
if hold is False:
pylab.clf()
pylab.hist(self._df.loc[:,'snr_A'], alpha=alpha, label="A", bins=bins)
pylab.hist(self._df.loc[:,'snr_C'], alpha=alpha, label="C", bins=bins)
pylab.hist(self._df.loc[:,'snr_G'], alpha=alpha, label="G", bins=bins)
pylab.hist(self._df.loc[:,'snr_T'], alpha=alpha, label="T", bins=bins)
pylab.legend()
pylab.xlabel(xlabel, fontsize=fontsize)
pylab.ylabel(ylabel, fontsize=fontsize)
if grid is True:
pylab.grid(True)
def plot_read_length(self):
"""Plot occurences of aligned read lengths
.. plot::
:include-source:
from sequana import sequana_data, BAM
b = BAM(sequana_data("test.bam"))
b.plot_read_length()
"""
X, Y = self._get_read_length()
pylab.plot(X, Y,
label="min length:{}; max length:{}".format(min(X), max(X)))
pylab.grid()
pylab.xlabel("Read length", fontsize=16)
pylab.legend()
def plot_read_length(self):
"""Plot occurences of aligned read lengths
.. plot::
:include-source:
from sequana import sequana_data, BAM
b = BAM(sequana_data("test.bam"))
b.plot_read_length()
"""
X, Y = self._get_read_length()
pylab.plot(X, Y,
label="min length:{}; max length:{}".format(min(X), max(X)))
pylab.grid()
pylab.xlabel("Read length", fontsize=16)
pylab.legend()
def barplot_count_ORF_CDS_by_frame(self, alpha=0.5, bins=40,
xlabel="Frame", ylabel="#", bar_width=0.35):
if self._ORF_pos is None:
self._find_ORF_CDS()
# number of ORF and CDS found by frame
frames = [-3, -2, -1, 1, 2, 3]
nb_res_ORF = []
nb_res_CDS = []
for fr in frames:
nb_res_ORF.append(self._ORF_pos[self._ORF_pos["frame"] == fr]["len_ORF"].dropna().shape[0])
nb_res_CDS.append(self._ORF_pos[self._ORF_pos["frame"] == fr]["len_CDS"].dropna().shape[0])
pylab.bar(np.array(frames)-(bar_width/2), nb_res_ORF, bar_width, alpha=alpha, label="ORF N = %d" %sum(nb_res_ORF))
pylab.bar(np.array(frames)+(bar_width/2), nb_res_CDS, bar_width, alpha=alpha, label="CDS N = %d" %sum(nb_res_CDS))
pylab.xlabel(xlabel)
pylab.ylabel(ylabel)
pylab.legend(loc=1)
pylab.title("Number of ORF and CDS by frame")
def hist_average_quality(self, fontsize=16):
hq_qv = [
mean([phred.ascii_to_quality(X) for X in read['quality'].decode()])
for read in iso.hq_sequence
]
lq_qv = [
mean([phred.ascii_to_quality(X) for X in read['quality'].decode()])
for read in iso.lq_sequence
]
Y1, X = numpy.histogram(hq_qv, bins=range(0, 94))
Y2, X = numpy.histogram(lq_qv, bins=range(0, 94))
pylab.bar(X[1:], Y1, width=1, label="HQ")
pylab.bar(X[1:], Y2, bottom=Y1, width=1, label="LQ")
pylab.xlabel("Isoform average QV")
pylab.ylabel("# Isoform")
pylab.legend(fontsize=fontsize)
def plot_alignment(self,
motif,
window=200,
global_th=10,
title=None,
legend=True,
legend_fontsize=11):
"""
plot alignments that match the motif.
"""
df = self._get_aligments(motif=motif,
window=window,
global_th=global_th)
print("Found {} hits".format(len(df)))
bam = BAM(self.bamfile)
pylab.clf()
count = 0
for aln in bam:
if aln.query_name in df.query_name.values:
seq = aln.query_sequence
if seq:
count += 1
X1 = [
seq[i:i + window].count(motif) for i in range(len(seq))
]
pylab.plot(range(aln.reference_start,
aln.reference_start + len(seq)),
X1,
label=aln.query_name)
max_theo = int(1.2 * window / len(motif))
pylab.ylim([0, max_theo])
if legend and count < 15:
pylab.legend(fontsize=legend_fontsize)
if title:
pylab.title(title, fontsize=16)
return df
def plot_scores(self, filename=None, savefig=False):
# scores
from pylab import log10
pylab.clf()
pylab.plot(log10(self.df.query('score>540')['rep1_signal']),
log10(self.df.query('score>540')['rep2_signal']),
'ko',
alpha=0.5,
label='<0.05 IDR')
pylab.plot(log10(self.df.query('score<540')['rep1_signal']),
log10(self.df.query('score<540')['rep2_signal']),
'ro',
alpha=0.5,
label='>=0.05 IDR')
N = pylab.ylim()[1]
pylab.plot([0, N], [0, N], color='blue', alpha=0.5, ls='--')
pylab.xlabel("Rep1 log10 score")
pylab.ylabel("Rep2 log10 score")
pylab.legend(loc='lower right')
if savefig:
pylab.savefig(filename)
def barplot(self, filename="lane{}_status.png", lanes=None):
df = self.get_data_reads()
if lanes is None:
lanes = df.lane.unique()
for lane in lanes:
pylab.clf()
query = "lane==@lane and name!='Undetermined'"
counts = df.query(query)['count']
total = counts.sum()
L = len(counts)
query = "lane==@lane and name=='Undetermined'"
under = df.query(query)['count'].sum()
if total > 0:
pylab.bar(range(L), counts, color="b", label="reads")
if total == 0:
color = "red"
else:
if 100 * under / total < 20:
color = "green"
elif 100 * under / total < 50:
color = "orange"
else:
color = "red"
pylab.bar(range(L, L + 1),
under,
color=color,
label="undetermined")
pylab.xticks([])
pylab.ylabel("Number of reads")
try:
pylab.legend(loc="lower left")
except:
pass
pylab.title("Lane {}".format(lane))
pylab.savefig(filename.format(lane), dpi=200)
def plot_bar_grouped(self, normalise=False, ncol=2, N=None):
"""
:param normalise:
:param ncol: columns in the legend
"""
if N is not None:
N = np.array(N)
else:
N = np.array([len(x) for x in self.rawdata])
dd = pd.DataFrame(self.sirv).T
if normalise:
dd = dd/ (N/max(N))
dd.columns = self.labels
dd.plot(kind="bar")
pylab.xlabel("")
pylab.legend(self.labels, ncol=ncol)
pylab.tight_layout()
return dd
def run(self, bins=50, xmin=0, xmax=30000, step=1000, burn=1000,alpha=1,output_filename=None):
# compute histogram of the input reads once for all to be used
# in the target_distribution method
self.bins = bins
self.Y, self.X = np.histogram(self.bam.df.read_length, bins=bins, density=True)
lengths = self.bam_simul.df.read_length.values
self.tokeep = []
vec = []
x = self.bam.df.read_length.mean()
for i in range(self.bam_simul.df.shape[0]):
can = lengths[i]
aprob = min([alpha,self.target_distribution(can)/self.target_distribution(x)])
#acceptance probability
u = pylab.uniform(0,1)
if u < aprob:
x = can
vec.append(x)
self.tokeep.append(True)
else:
self.tokeep.append(False)
#plotting the results:
#theoretical curve
x = pylab.arange(xmin, xmax, step)
y = self.target_distribution(x)
pylab.subplot(211)
pylab.title('Metropolis-Hastings')
pylab.plot(vec)
pylab.subplot(212)
pylab.hist(vec[burn:], bins=bins, density=1)
pylab.plot(x,y,'r-')
pylab.ylabel('Frequency')
pylab.xlabel('x')
pylab.legend(('PDF','Samples'))
if output_filename is not None:
self.bam_simul.filter_bool(output_filename, self.tokeep)
def plot_bar_grouped(self, normalise=False, ncol=2, N=None):
"""
:param normalise:
:param ncol: columns in the legend
"""
if N is not None:
N = np.array(N)
else:
N = np.array([len(x) for x in self.rawdata])
dd = pd.DataFrame(self.sirv).T
if normalise:
dd = dd / (N / max(N))
dd.columns = self.labels
dd.plot(kind="bar")
pylab.xlabel("")
pylab.legend(self.labels, ncol=ncol)
pylab.tight_layout()
return dd
def _do_legend(self, figure, color_dict, bbox_to_anchor):
if color_dict:
patches = [
mpatches.Patch(color=c, label=l)
for l, c in color_dict.items()
]
legend = pylab.legend(
loc="upper center",
handles=patches,
bbox_to_anchor=bbox_to_anchor,
frameon=True,
title="Sample groups",
)
figure.add_artist(legend)
