def plot(self,
color_line='r',
bgcolor='grey',
color='yellow',
lw=4,
hold=False,
ax=None):
xmax = self.xmax + 1
if ax:
pylab.sca(ax)
pylab.fill_between([0, xmax], [0, 0], [20, 20], color='red', alpha=0.3)
pylab.fill_between([0, xmax], [20, 20], [30, 30],
color='orange',
alpha=0.3)
pylab.fill_between([0, xmax], [30, 30], [41, 41],
color='green',
alpha=0.3)
if self.X is None:
X = range(1, self.xmax + 1)
pylab.fill_between(X,
self.df.mean() + self.df.std(),
self.df.mean() - self.df.std(),
color=color,
interpolate=False)
pylab.plot(X, self.df.mean(), color=color_line, lw=lw)
pylab.ylim([0, 41])
pylab.xlim([0, self.xmax + 1])
pylab.title("Quality scores across all bases")
pylab.xlabel("Position in read (bp)")
pylab.ylabel("Quality")
pylab.grid(axis='x')
def plot_contig_length_vs_GC(self, alpha=0.5):
pylab.plot(self.df["length"], self.df['GC'], "o", alpha=alpha)
pylab.xlabel("contig length (bp)")
pylab.ylabel("GC (%)")
pylab.grid(True)
pylab.ylim([0, 100])
pylab.xlim(0, max(self.df['length']) + 10)
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 imshow_qualities(self):
"""Qualities
::
from sequana import sequana_data
from sequana import FastQC
filename = sequana_data("test.fastq", "testing")
qc = FastQC(filename)
qc.imshow_qualities()
from pylab import tight_layout; tight_layout()
"""
tiles = self._get_tile_info()
d = defaultdict(list)
for tile, seq in zip(tiles['tiles'], self.qualities):
d[tile].append(seq)
self.data_imqual = [pd.DataFrame(d[key]).mean().values for key in sorted(d.keys())]
from biokit.viz import Imshow
im = Imshow(self.data_imqual)
im.plot(xticks_on=False, yticks_on=False, origin='lower')
pylab.title("Quality per tile", fontsize=self.fontsize)
pylab.xlabel("Position in read (bp)")
pylab.ylabel("tile number")
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 plot_bar_flags(self, logy=True, fontsize=16, filename=None):
"""Plot an histogram of the flags contained in the BAM
.. plot::
:include-source:
from sequana import BAM, sequana_data
b = BAM(sequana_data('test.bam', "testing"))
b.plot_bar_flags()
.. seealso:: :class:`SAMFlags` for meaning of each flag
"""
df = self.get_flags_as_df()
df = df.sum()
pylab.clf()
if logy is True:
barplot = df.plot(kind='bar', logy=logy, grid=True)
else:
barplot = df.plot(kind='bar', grid=True)
pylab.xlabel("flags", fontsize=fontsize)
pylab.ylabel("count", fontsize=fontsize)
pylab.tight_layout()
if filename:
pylab.savefig(filename)
return barplot
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_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 hist_read_length_consensus_isoform(self, mode="all", bins=80, rwidth=0.8,
align="left", fontsize=16, edgecolor="k", **kwargs):
"""
mode can be all, lq, hq
"""
pylab.clf()
L1 = [len(read['sequence']) for read in self.lq_sequence]
L2 = [len(read['sequence']) for read in self.hq_sequence]
if mode == "all":
L = L1 + L2
elif mode == "lq":
L = L1
else:
L = L2
Y, X, _ = pylab.hist(L, bins=bins, rwidth=rwidth, align=align,
ec=edgecolor, **kwargs)
pylab.gca().set_ylim(bottom=0)
pylab.gca().set_xlim(left=0)
pylab.xlabel("Read length", fontsize=fontsize)
pylab.ylabel("Number of reads", fontsize=fontsize)
pylab.grid()
ax_twin = pylab.gca().twinx()
shift = (X[1] - X[0]) / 2
ax_twin.plot(X[0:-1]- shift, len(L) - pylab.cumsum(Y), "k")
ax_twin.set_ylim(bottom=0)
pylab.ylabel("CDF", fontsize=fontsize)
pylab.title("Read length of Consensus isoforms reads")
def plot_bar_flags(self, logy=True, fontsize=16, filename=None):
"""Plot an histogram of the flags contained in the BAM
.. plot::
:include-source:
from sequana import BAM, sequana_data
b = BAM(sequana_data('test.bam', "testing"))
b.plot_bar_flags()
.. seealso:: :class:`SAMFlags` for meaning of each flag
"""
df = self.get_flags_as_df()
df = df.sum()
pylab.clf()
if logy is True:
barplot = df.plot(kind='bar', logy=logy, grid=True)
else:
barplot = df.plot(kind='bar', grid=True)
pylab.xlabel("flags", fontsize=fontsize)
pylab.ylabel("count", fontsize=fontsize)
pylab.tight_layout()
if filename:
pylab.savefig(filename)
return barplot
def hist_read_length_consensus_isoform(self, mode="all", bins=80, rwidth=0.8,
align="left", fontsize=16, edgecolor="k", **kwargs):
"""
mode can be all, lq, hq
"""
pylab.clf()
L1 = [len(read['sequence']) for read in self.lq_sequence]
L2 = [len(read['sequence']) for read in self.hq_sequence]
if mode == "all":
L = L1 + L2
elif mode == "lq":
L = L1
else:
L = L2
Y, X, _ = pylab.hist(L, bins=bins, rwidth=rwidth, align=align,
ec=edgecolor, **kwargs)
pylab.gca().set_ylim(bottom=0)
pylab.gca().set_xlim(left=0)
pylab.xlabel("Read length", fontsize=fontsize)
pylab.ylabel("Number of reads", fontsize=fontsize)
pylab.grid()
ax_twin = pylab.gca().twinx()
shift = (X[1] - X[0]) / 2
ax_twin.plot(X[0:-1]- shift, len(L) - pylab.cumsum(Y), "k")
ax_twin.set_ylim(bottom=0)
pylab.ylabel("CDF", fontsize=fontsize)
pylab.title("Read length of Consensus isoforms reads")
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_bar_mapq(self, fontsize=16, filename=None):
"""Plots bar plots of the MAPQ (quality) of alignments
.. plot::
:include-source:
from sequana import BAM, sequana_data
b = BAM(sequana_data('test.bam', "testing"))
b.plot_bar_mapq()
"""
df = self.get_mapq_as_df()
df.plot(kind='hist',
bins=range(0,
df.max().values[0] + 1),
legend=False,
grid=True,
logy=True)
pylab.xlabel("MAPQ", fontsize=fontsize)
pylab.ylabel("Count", fontsize=fontsize)
try:
# This may raise issue on MAC platforms
pylab.tight_layout()
except:
pass
if filename:
pylab.savefig(filename)
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 hist_length_repeats(self,
bins=20,
alpha=0.5,
hold=False,
fontsize=12,
grid=True,
title="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 hold is False:
pylab.clf()
pylab.hist(self._list_len_repeats, alpha=alpha, bins=bins)
pylab.title(title)
pylab.xlabel(xlabel, fontsize=fontsize)
pylab.ylabel(ylabel, fontsize=fontsize)
if grid is True:
pylab.grid(True)
def scatterplot(self, enrich, cutoff=0.05, nmax=10, gene_set_size=[]):
df = self._get_final_df(enrich.results, cutoff=cutoff, nmax=nmax)
pylab.clf()
pylab.scatter(-pylab.log10(df['Adjusted P-value']),
range(len(df)),
s=10 * df['size'],
c=df['size'])
pylab.xlabel("Odd ratio")
pylab.ylabel("Gene sets")
pylab.yticks(range(len(df)), df.name)
a, b = pylab.xlim()
pylab.xlim([0, b])
pylab.grid(True)
ax = pylab.gca()
M = max(df['size'])
if M > 100:
l1, l2, l3 = "10", "100", str(M)
else:
l1, l2, l3 = str(round(M / 3)), str(round(M * 2 / 3)), str(M)
handles = [
pylab.Line2D([0], [0], marker="o", markersize=5, label=l1, ls=""),
pylab.Line2D([0], [0], marker="o", markersize=10, label=l2, ls=""),
pylab.Line2D([0], [0], marker="o", markersize=15, label=l3, ls="")
]
ax.legend(handles=handles, loc="upper left", title="gene-set size")
pylab.axvline(1.3, lw=2, ls="--", color="r")
pylab.tight_layout()
ax = pylab.colorbar(pylab.gci())
return df
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 get_max_gc_correlation(self, reference):
"""Plot correlation between coverage and GC content by varying the GC window
The GC content uses a moving window of size W. This parameter affects
the correlation bewteen coverage and GC. This function find the
*optimal* window length.
"""
pylab.clf()
corrs = []
wss = []
def func(params):
ws = int(round(params[0]))
if ws < 10:
return 0
self.bed.compute_gc_content(reference, ws)
corr = self.get_gc_correlation()
corrs.append(corr)
wss.append(ws)
return corr
from scipy.optimize import fmin
res = fmin(func, 100, xtol=1, disp=False) # guess is 200
pylab.plot(wss, corrs, "o")
pylab.xlabel("GC window size")
pylab.ylabel("Correlation")
pylab.grid()
return res[0]
def hist_entropy(self, bins=50):
"""Histogram of the entropy of all found repeats
"""
self.df.entropy.hist(bins=bins)
pylab.xlabel("Entropy")
pylab.ylabel("#")
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 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 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 histogram_sequence_lengths(self, logy=True):
"""Histogram sequence lengths
.. plot::
:include-source:
from sequana import sequana_data
from sequana import FastQC
filename = sequana_data("test.fastq", "testing")
qc = FastQC(filename)
qc.histogram_sequence_lengths()
"""
data = [len(x) for x in self.sequences]
bary, barx = np.histogram(data, bins=range(max(data)+1))
# get rid of zeros to avoid warnings
bx = [x for x,y in zip(barx, bary) if y!=0]
by = [y for x,y in zip(barx, bary) if y!=0]
if logy:
pylab.bar(bx, pylab.log10(by))
else:
pylab.bar(bx, by)
pylab.xlim([1,max(data)+1])
pylab.grid(True)
pylab.xlabel("position (bp)", fontsize=self.fontsize)
pylab.ylabel("Count (log scale)", fontsize=self.fontsize)
def plot_count_per_sample(self, fontsize=12, sample_list=None):
""""Number of mapped reads per sample. Each color for each replicate
.. plot::
:include-source:
from sequana.rnadiff import RNADiffResults
from sequana import sequana_data
r = RNADiffResults(sequana_data("rnadiff/rnadiff_onecond_1"))
r.plot_count_per_sample()
"""
sample_names = self.sample_names
N = len(sample_names)
dd = self.df[sample_names].sum()
pylab.clf()
colors = []
for sample in self.sample_names:
colors.append(self.colors[self.get_cond_from_sample(sample)])
pylab.bar(range(N), (dd/1000000).values,
color=colors, alpha=1,
zorder=10, lw=1, ec="k", width=0.9)
pylab.xlabel("Samples", fontsize=fontsize)
pylab.ylabel("Total read count (millions)", fontsize=fontsize)
pylab.grid(True, zorder=0)
pylab.title("Total read count per sample", fontsize=fontsize)
pylab.xticks(range(N), self.sample_names)
def hist_contig_length(self, bins=30, fontsize=16):
pylab.clf()
pylab.hist(self.df.length, lw=1, ec="k", bins=bins)
pylab.grid()
pylab.xlabel("Contig length", fontsize=fontsize)
pylab.ylabel("#", fontsize=fontsize)
pylab.title("Distribution {} contigs".format(len(self.df)))
def plot_volcano(self):
"""
.. plot::
:include-source:
from sequana.rnadiff import RNADiffResults
from sequana import sequana_data
r = RNADiffResults(sequana_data("rnadiff/rnadiff_onecond_1"))
r.plot_volcano()
"""
d1 = self.df.query("padj>0.05")
d2 = self.df.query("padj<=0.05")
fig = pylab.figure()
pylab.plot(d1.log2FoldChange, -np.log10(d1.padj), marker="o",
alpha=0.5, color="r", lw=0)
pylab.plot(d2.log2FoldChange, -np.log10(d2.padj), marker="o",
alpha=0.5, color="k", lw=0)
pylab.grid(True)
pylab.xlabel("fold change")
pylab.ylabel("log10 adjusted p-value")
m1 = abs(min(self.df.log2FoldChange))
m2 = max(self.df.log2FoldChange)
limit = max(m1,m2)
pylab.xlim([-limit, limit])
y1,y2 = pylab.ylim()
pylab.ylim([0,y2])
pylab.axhline(-np.log10(0.05), lw=2, ls="--", color="r", label="pvalue threshold (0.05)")
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 plot_idr_vs_peaks(self, filename=None, savefig=False):
# global_idr is actually -log10(idr)
pylab.clf()
X1 = pylab.linspace(0, self.threshold, 100)
X2 = pylab.linspace(self.threshold, 1, 100)
# convert global idr to proba
df1 = self.df.query("idr<@self.threshold")
df2 = self.df.query("idr>[email protected]")
pylab.plot([sum(df1['idr'] < x) for x in X1], X1, '-', color='r', lw=2)
shift = len(df1)
pylab.plot([shift + sum(df2['idr'] < x) for x in X2],
X2,
"-",
color='k',
lw=2)
pylab.xlabel('Number of significant peaks')
pylab.ylabel('IDR')
pylab.axhline(0.05, color='b', ls='--')
pylab.axvline(self.N_significant_peaks, color='b', ls='--')
if savefig:
pylab.savefig(filename)
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 hist_plot_contig_length(self, bins=40, fontsize=16):
"""Plot distribution of contig lengths"""
L = len(self.fasta.sequences)
pylab.hist(self.fasta.lengths, lw=1, ec="k", bins=bins)
pylab.grid()
pylab.xlabel("Contig length", fontsize=fontsize)
pylab.ylabel("#", fontsize=fontsize)
pylab.title("Distribution {} contigs".format(L))
def plot_subreads_histogram(self, bins=10, fontsize=12):
self.df_barcoded['Subreads'].hist(bins=bins, ec="k", rwidth=0.8)
pylab.xlabel("Number of subreads", fontsize=fontsize)
pylab.ylabel("Number of Barcoded Samples", fontsize=fontsize)
try:
pylab.tight_layout()
except:
pass
def plot_pvalue_hist(self, bins=60, fontsize=16, rotation=0):
pylab.hist(self.df.pvalue.dropna(), bins=bins, ec="k")
pylab.grid(True)
pylab.xlabel("raw p-value", fontsize=fontsize)
pylab.ylabel("Occurences", fontsize=fontsize)
try:
pylab.tight_layout()
except:
pass
def plot_padj_hist(self, bins=60, fontsize=16):
pylab.hist(self.df.padj.dropna(), bins=bins, ec="k")
pylab.grid(True)
pylab.xlabel("Adjusted p-value", fontsize=fontsize)
pylab.ylabel("Occurences", fontsize=fontsize)
try:
pylab.tight_layout()
except:
pass
def scatter_length_cov_gc(self, min_length=200, min_cov=10):
pylab.clf()
pylab.scatter(self.df.length, self.df['cov'], c=self.df.GC)
pylab.loglog()
pylab.axvline(min_length, lw=2, c="r", ls='--')
pylab.axhline(min_cov, lw=2, c="r", ls='--')
pylab.xlabel("contig length")
pylab.ylabel("contig coverage")
pylab.colorbar(label="GC")
pylab.grid(True)
def plot(self, fontsize=16):
"""plot quality versus base position"""
pylab.plot(self.quality, label="offset: %s" % self.offset)
pylab.xlabel('base position', fontsize=fontsize)
pylab.ylabel('Quality per base', fontsize=fontsize)
pylab.grid(True)
# ylim set autoscale to off so if we want to call this function several
# times, we must reset autoscale to on before calling ylim
pylab.autoscale()
limits = pylab.ylim()
pylab.ylim(max(0,limits[0]-1), limits[1]+1)
def boxplot_mapq_concordance(self):
# method can only be bwa for now
assert self.method == "bwa"
data = self._get_data()
df = pd.DataFrame(data, columns=["mapq", "length", "concordance"])
pylab.clf()
pylab.boxplot([df[df.mapq == i]['concordance'] for i in range(1,61)])
pylab.xlabel("mapq")
pylab.ylabel("concordance")
pylab.grid()
tt = [10,20,30,40,50,60]
pylab.xticks(tt, tt)
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 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_coverage(self, bins=100):
"""
.. plot::
:include-source:
from sequana import sequana_data, BAM
b = BAM(sequana_data("measles.fa.sorted.bam"))
b.hist_coverage()
"""
try: self.coverage
except: self._set_coverage()
pylab.hist(self.coverage, bins=bins)
pylab.xlabel("Coverage")
pylab.ylabel("Number of mapped bases")
pylab.grid()
def hist_length_repeats(self, bins=20, alpha=0.5, hold=False,
fontsize=12, grid=True, title="Repeat length",
xlabel="Repeat length", ylabel="#", logy=True):
"""Plots histogram of the repeat lengths
"""
# check that user has set a threshold
if hold is False:
pylab.clf()
pylab.hist(self.list_len_repeats, alpha=alpha, bins=bins)
pylab.title(title)
pylab.xlabel(xlabel, fontsize=fontsize)
pylab.ylabel(ylabel, fontsize=fontsize)
if grid is True:
pylab.grid(True)
if logy:
pylab.semilogy()
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_bar_mapq(self, fontsize=16, filename=None, ):
"""Plots bar plots of the MAPQ (quality) of alignments
.. plot::
:include-source:
from sequana import BAM, sequana_data
b = BAM(sequana_data('test.bam', "testing"))
b.plot_bar_mapq()
"""
df = self.get_mapq_as_df()
df.plot(kind='hist', bins=range(0,df.max().values[0]+1), legend=False,
grid=True, logy=True)
pylab.xlabel("MAPQ", fontsize=fontsize)
pylab.ylabel("Count", fontsize=fontsize)
pylab.tight_layout()
if filename:
pylab.savefig(filename)
def pie_plot(self, filename=None, hold=False):
"""Plot PIE plot of the status (complete / fragment / missed)
.. plot::
:include-source:
from sequana import BUSCO, sequana_data
b = BUSCO(sequana_data("test_busco_full_table.tsv"))
b.pie_plot()
"""
if hold is False:
pylab.clf()
self.df.groupby('Status').count()['# Busco id'].plot(kind="pie")
pylab.ylabel("")
#pylab.title("Distribution Complete/Fragmented/Missing")
#pylab.legend()
if filename:
pylab.savefig(filename)
def hist_GC(self, bins=50, alpha=0.5, hold=False, fontsize=12,
grid=True, xlabel="GC %", ylabel="#", label="",title=None):
"""Plot histogram GC content
:param int bins: binning for the histogram
:param float alpha: transparency of the histograms
:param bool hold:
:param int fontsize: fontsize of the x and y labels and title.
:param bool grid: add grid or not
:param str xlabel:
:param str ylabel:
:param str label: label of the histogram (for the legend)
: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_GC()
"""
mean_GC = np.mean(self.df.loc[:,'GC_content'])
# set title if needed
if title is None:
title = "GC %% \n Mean GC : %.2f" %(mean_GC)
# histogram GC percent
if hold is False:
pylab.clf()
pylab.hist(self.df.loc[:,'GC_content'], bins=bins,
alpha=alpha, label=label + ", mean : " + str(round(mean_GC, 2))
+ ", N : " + str(len(self)))
pylab.xlabel(xlabel, fontsize=fontsize)
pylab.ylabel(ylabel, fontsize=fontsize)
pylab.title(title, fontsize=fontsize)
if grid is True:
pylab.grid(True)
pylab.xlim([0, 100])
try: pylab.tight_layout()
except:pass
def plot_acgt_content(self, stacked=False):
"""Plot histogram of GC content
.. plot::
:include-source:
from sequana import sequana_data
from sequana import FastQC
filename = sequana_data("test.fastq", "testing")
qc = FastQC(filename)
qc.plot_acgt_content()
"""
df = self.get_actg_content()
if stacked is True:
df.plot.bar(stacked=True)
else:
df.plot()
pylab.grid(True)
pylab.xlabel("position (bp)", fontsize=self.fontsize)
pylab.ylabel("percent", fontsize=self.fontsize)
def hist_nb_passes(self, bins=None, alpha=0.5, hold=False, fontsize=12,
grid=True, xlabel="Number of ZMW passes", logy=True,
ylabel="#", label="", title="Number of ZMW passes"):
"""Plot histogram of number of reads per ZMW (number of passes)
:param float alpha: transparency of the histograms
:param bool hold:
:param int fontsize:
:param bool grid:
:param str xlabel:
:param str ylabel:
:param bool logy: use log scale on the y axis (default to True)
:param str label: label of the histogram (for the legend)
: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_nb_passes()
"""
max_nb_pass = self.df.nb_passes.max()
if bins is None:
k = range(1, max_nb_pass+1)
# histogram nb passes
if hold is False:
pylab.clf()
pylab.hist(self.df.nb_passes, bins=bins, alpha=alpha,
label=label, log=logy, width=1)
if len(k) < 5:
pylab.xticks(range(6), range(6))
pylab.xlabel(xlabel, fontsize=fontsize)
pylab.ylabel(ylabel, fontsize=fontsize)
pylab.title(title, fontsize=fontsize)
if grid is True:
pylab.grid(True)
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(self, bins=80, rwidth=0.8, **kwargs):
pylab.clf()
Y, X, _ = pylab.hist(self.data, bins=bins, rwidth=rwidth, **kwargs)
pylab.xlabel(self.xlabel, fontsize=self.fontsize)
pylab.ylabel(self.ylabel, fontsize=self.fontsize)
"""self.Y = Y
self.X = X
ax_twin = pylab.gca().twinx()
shift = (X[1] - X[0]) / 2
ax_twin.plot(X[0:-1]- shift, len(self.data) - pylab.cumsum(Y), "k")
ax_twin.set_ylim(bottom=0)
pylab.ylabel("CDF", fontsize=self.fontsize)
"""
pylab.grid(self.grid)
pylab.title(self.title)
try: pylab.tight_layout()
except:pass
def plot_GC_read_len(self, hold=False, fontsize=12, bins=[200, 60],
grid=True, xlabel="GC %", ylabel="#", cmap="BrBG"):
"""Plot GC content versus read length
:param bool hold:
:param int fontsize: for x and y labels and title
:param bins: a integer or tuple of 2 integers to specify
the binning of the x and y 2D histogram.
:param bool grid:
:param str xlabel:
:param str ylabel:
.. plot::
:include-source:
from sequana.pacbio import PacbioSubreads
from sequana import sequana_data
b = PacbioSubreads(sequana_data("test_pacbio_subreads.bam"))
b.plot_GC_read_len(bins=[10, 10])
"""
mean_len = np.mean(self.df.loc[:,'read_length'])
mean_GC = np.mean(self.df.loc[:,'GC_content'])
if hold is False:
pylab.clf()
data = self.df.loc[:,['read_length','GC_content']].dropna()
h = biokit.viz.hist2d.Hist2D(data)
res = h.plot(bins=bins, contour=False, norm='log', Nlevels=6, cmap=cmap)
pylab.xlabel("Read length", fontsize=fontsize)
pylab.ylabel("GC %", fontsize=fontsize)
pylab.title("GC %% vs length \n Mean length : %.2f , Mean GC : %.2f" %
(mean_len, mean_GC), fontsize=fontsize)
pylab.ylim([0, 100])
if grid is True:
pylab.grid(True)
def plot(self, kind="pie", cmap="copper", threshold=1, radius=0.9,
textcolor="red", **kargs):
"""A simple non-interactive plot of taxons
:return: None if no taxon were found and a dataframe otherwise
A Krona Javascript output is also available in :meth:`kraken_to_krona`
.. plot::
:include-source:
from sequana import KrakenResults, sequana_data
test_file = sequana_data("test_kraken.out", "testing")
k = KrakenResults(test_file)
df = k.plot(kind='pie')
.. seealso:: to generate the data see :class:`KrakenPipeline`
or the standalone application **sequana_taxonomy**.
"""
if len(self._df) == 0:
return
if self._data_created == False:
status = self.kraken_to_krona()
if kind not in ['barh', 'pie']:
logger.error('kind parameter: Only barh and pie are supported')
return
# This may have already been called but maybe not. This is not time
# consuming, so we call it again here
if len(self.taxons.index) == 0:
return None
df = self.get_taxonomy_biokit(list(self.taxons.index))
df.ix[-1] = ["Unclassified"] * 8
data = self.taxons.copy()
data.ix[-1] = self.unclassified
data = data/data.sum()*100
assert threshold > 0 and threshold < 100
others = data[data<threshold].sum()
data = data[data>threshold]
names = df.ix[data.index]['name']
data.index = names.values
data.ix['others'] = others
try:
data.sort_values(inplace=True)
except:
data.sort(inplace=True)
# text may be long so, let us increase the figsize a little bit
pylab.figure(figsize=(10,8))
pylab.clf()
if kind == "pie":
ax = data.plot(kind=kind, cmap=cmap, autopct='%1.1f%%',
radius=radius, **kargs)
pylab.ylabel(" ")
for text in ax.texts:
# large, x-small, small, None, x-large, medium, xx-small,
# smaller, xx-large, larger
text.set_size("small")
text.set_color(textcolor)
for wedge in ax.patches:
wedge.set_linewidth(1)
wedge.set_edgecolor("k")
self.ax = ax
elif kind == "barh":
ax = data.plot(kind=kind, **kargs)
pylab.xlabel(" percentage ")
return data
def hist_passes(self, maxp=50, fontsize=16):
passes = self.df.nb_passes.copy()
passes.clip_upper(maxp).hist(bins=maxp)
pylab.xlim([0, maxp])
pylab.ylabel("# count", fontsize=fontsize)
pylab.xlabel("Passes (max {})".format(maxp), fontsize=fontsize)