def size_dist(inputs, paths_in, paths_out):
files = inputs['files']
path_figure = paths_out['path_figures']
plot_num = 0
sns.set_style("white")
plt.figure(figsize=(5, 3))
naming = ''
for fname in files:
naming += fname + '_'
path_analysis = paths_out['path_analysis'] + fname + '/readQC/'
data = ribo_util.unPickle(path_analysis + 'read_distribution')
df = ribo_util.dict_to_df(data, 'Length', 'fraction of total')
plt.plot(df, label=fname)
plt.title('Size Distribution')
plt.xlabel("Read Length")
plt.ylabel("Percent of Reads")
plt.legend(bbox_to_anchor=(1.05, 1), loc=2, borderaxespad=0.)
plt.savefig(path_figure + 'Comparison/size_dist' + '/sizedist_' + naming +
'.pdf',
dpi=400,
bbox_inches="tight")
plt.show()
for fname in files:
sns.set_style("white")
plt.figure(figsize=(5, 3))
path_analysis = paths_out['path_analysis'] + fname + '/readQC/'
data = ribo_util.unPickle(path_analysis + 'read_distribution')
df = ribo_util.dict_to_df(data, 'Length', 'fraction of total')
plt.plot(df, label=fname)
plt.title('Size Distribution')
plt.xlabel("Read Length")
plt.ylabel("Percent of Reads")
plt.legend(loc='upper right')
plt.savefig(path_figure + fname + '/sizedist.pdf',
dpi=400,
bbox_inches="tight")
size_plot_csv = pd.DataFrame(df)
size_plot_csv.to_csv(path_analysis + 'size_plot_values.csv')
plt.gcf().clear()
def read_comp(inputs, paths_in, paths_out):
files = inputs['files']
path_figure = paths_out['path_figures']
plot_num = 0
for fname in files:
plot_num = 0
plt.figure(figsize=(20, 2.5))
for nucleotide in ['G', 'C', 'A', 'U']:
plot_num += 1
path_analysis = paths_out['path_analysis'] + fname + '/readQC/'
data = ribo_util.unPickle(path_analysis + 'comp_' + nucleotide)
df = ribo_util.heatmapdict_to_df(data, 'Length', 'Position',
'composition')
plt.subplot(1, 4, plot_num)
plot = sns.heatmap(df, cmap="RdBu_r", vmin=0, vmax=50)
plt.setp(plot.get_xticklabels(), visible=False)
plt.setp(plot.get_xticklabels()[0::5], visible=True)
plt.setp(plot.get_yticklabels(), visible=False)
plt.setp(plot.get_yticklabels()[0::4], visible=True)
plt.title(fname + ' ' + nucleotide)
plt.savefig(path_figure + fname + '/read_composition.pdf',
dpi=400,
bbox_inches="tight")
plt.show()
def plot_alignment_allocation(inputs, paths_in, paths_out):
files = inputs['files']
analysis_path = paths_out['path_analysis_log']
path_figure = paths_out['path_figures']
data = {}
fnames = []
samples = 0
samples_list = []
for fname in files:
fnames.append(fname)
samples_list.append(samples)
samples += 1
analysis_log = analysis_path + fname + '/' + fname
analysis_log = ribo_util.unPickle(analysis_log)
raw_data = analysis_log['ribo_density']['analysis_breakdown']
total_reads = raw_data.pop('Total Reads', None)
#total_reads = raw_data.pop('Reads Filtered', None)
for key in raw_data.keys():
if not key in data:
data[key] = []
value = raw_data[key]
data[key].append(value)
legend = raw_data.keys()
legend = sorted(legend, key=str.lower)
colors = ['#b2182b','#fddbc7','#e0e0e0','#bababa','#878787','#4d4d4d']
bottom = [0]*samples
sns.set_style("white")
plt.figure(figsize=(.8 * samples,4))
for key in legend:
i = legend.index(key)
color = colors[i]
plt.bar(samples_list, data[key], bottom = bottom, color = color, edgecolor='white', width=.6)
bottom = [x + y for x, y in zip(bottom, data[key])]
plt.xticks(samples_list, fnames, fontweight='bold')
plt.title('Read Mapping')
plt.xlabel("Sample")
plt.ylabel("Reads")
plt.legend(legend, bbox_to_anchor=(1.05, 1), loc=2, borderaxespad=0.)
plt.savefig(path_figure + fname + '/read_allocation.pdf', dpi=400, bbox_inches="tight")
plt.show()
def plot_asymmetry_comp(inputs, paths_in, paths_out, settings):
#sns.boxplot(data=iris, orient="h", palette="Set2")
files = inputs['files']
path_figure = paths_out['path_figures'] + 'Comparison/'
minlength = settings['minlength']
maxlength = settings['maxlength']
minlength_1 = str(minlength) + '_'
maxlength_1 = str(maxlength) + '_'
name_settings = minlength_1 + maxlength_1
data = {}
score_list = []
fname_list = []
plot_count = 0
fnames = ''
for fname in files:
fnames += fname
plot_count += 1
path_analysis = paths_out['path_analysis'] + fname + '/'
asymmetry_dict = ribo_util.unPickle(path_analysis + 'asymmetry')
df = pd.DataFrame(asymmetry_dict)
#display(df)
score = df['Score'].values.tolist()
for value in score:
fname_list.append(fname)
score_list.append(value)
data['file'] = fname_list
data['score'] = score_list
data = pd.DataFrame(data)
sns.set_style("white")
plot = plt.figure(figsize=(20, 10))
plot = sns.boxplot(x="score",
y='file',
data=data,
orient="h",
color="grey",
fliersize=0)
plot.axvline(linewidth=3, color='r')
plot.set(xlim=(-5, 5))
plot.tick_params(labelsize=18)
sns.despine(offset=5)
plt.savefig(path_figure + 'asymmetry/' + fnames + name_settings + '.pdf',
dpi=600,
bbox_inches="tight")
plt.show()
def plot_asymmetry(inputs, paths_in, paths_out, settings):
files = inputs['files']
path_figure = paths_out['path_figures']
minlength = settings['minlength']
maxlength = settings['maxlength']
minlength_1 = str(minlength) + '_'
maxlength_1 = str(maxlength) + '_'
name_settings = minlength_1 + maxlength_1
for fname in files:
path_analysis = paths_out['path_analysis'] + fname + '/'
asymmetry_dict = ribo_util.unPickle(path_analysis + 'asymmetry')
df = pd.DataFrame(asymmetry_dict)
#display(df)
sub = df.loc[df.Subgroup == "Subgroup"]
other = df.loc[df.Subgroup == "Other"]
plot2 = sns.jointplot(x="Score",
y="GeneLength",
data=df,
size=4,
color="black",
marker='.',
alpha=0,
xlim=(-3, 3))
plot2.x = other.Score
plot2.y = other.GeneLength
plot2.plot_joint(plt.scatter, marker='.', c='.6')
plot2.plot_joint(sns.kdeplot, n_levels=6, cmap="Greys")
plot2.x = sub.Score
plot2.y = sub.GeneLength
plot2.plot_joint(plt.scatter, marker='.', c='orangered', alpha=.6)
plot2.plot_joint(sns.kdeplot, n_levels=4, cmap="Reds")
plt.savefig(path_figure + fname + '/asymmetry' + name_settings +
'.pdf',
dpi=400,
bbox_inches="tight")
plt.show()
def plot_frame_comp(inputs, paths_in, paths_out, settings, settings_plot):
files = inputs['files']
minlength = settings['minlength']
maxlength = settings['maxlength']
threshold = settings['threshold']
minlength_1 = str(minlength) + '_'
maxlength_1 = str(maxlength) + '_'
threshold_1 = str(threshold) + '_'
name_settings = minlength_1 + maxlength_1 + threshold_1
framescore = []
for fname in files:
path_analysis = paths_out['path_analysis'] + fname + '/frame/'
frame_genome = ribo_util.unPickle(path_analysis + name_settings +
'frame_genome')
x_genome = [0, 1, 2]
frame_score = abs(frame_genome[0] - 40.3)
frame_score = frame_score / 59.7
framescore.append(frame_score)
print frame_score
print files, framescore
plt.figure(figsize=(30, 5))
plot1 = sns.barplot(x=files, y=framescore, color=".2")
plt.ylabel("Libraries")
plt.xlabel("Frame Score")
for p in plot1.patches:
height = p.get_height()
plot1.annotate("%.1f" % p.get_height(),
(p.get_x() + p.get_width() / 2., p.get_height() - 50),
ha='center',
va='center',
fontsize=10,
color='0',
xytext=(0, 20),
textcoords='offset points')
plt.show()
def plot_genelist(fnamelist, paths_in, paths_out, settings):
'''Takes 2 files and compares gene RPKM values'''
path_figure = paths_out['path_figures'] + 'Comparison/'
'''Load Settings'''
file1 = fnamelist[0]
file2 = fnamelist[1]
foldchange = settings['foldchange']
highlight_genes = settings['interesting']
data_transform = settings[
'data_transform'] # has to be 'None', 'log10', or 'log2'
'''Load Data'''
#genelist file from analysis
path_analysis1 = paths_out['path_analysis'] + fnamelist[0] + '/'
path_analysis2 = paths_out['path_analysis'] + fnamelist[1] + '/'
genelist_f1 = ribo_util.unPickle(path_analysis1 + 'genelist')
genelist_f2 = ribo_util.unPickle(path_analysis2 + 'genelist')
#define alias and RPKM as arrays
df1 = pd.DataFrame(genelist_f1)
alias = df1.Alias
RPKM_f1 = df1.RPKM
df2 = pd.DataFrame(genelist_f2)
RPKM_f2 = df2.RPKM
'''data arrays to plot'''
xydict = {}
x_list = []
y_list = []
folddict = {}
x_foldchange = []
y_foldchange = []
interestingdict = {}
x_interesting = []
y_interesting = []
'''iterate through gene data for manipulation (log transform and fold change calc)'''
for alias, xval, yval in itertools.izip(alias.values, RPKM_f1.values,
RPKM_f2.values):
# remove genes with RPKM = 0
if xval == 0 or yval == 0:
continue
# calculate foldchange prior to data transformation
foldchange_val = (yval / xval)
# transform data
if data_transform == 'None':
xval = xval
yval = yval
if data_transform == 'log2':
xval = math.log(xval, 2)
yval = math.log(yval, 2)
if data_transform == 'log10':
xval = math.log10(xval)
yval = math.log10(yval)
# append data to lists
x_list.append(xval)
y_list.append(yval)
if foldchange_val > foldchange or foldchange_val < (1 / foldchange):
x_foldchange.append(xval)
y_foldchange.append(yval)
if alias in settings['interesting']:
x_interesting.append(xval)
y_interesting.append(yval)
# convert lists to pd.dataframe
xydict[file1] = x_list
xydict[file2] = y_list
xy_df = pd.DataFrame(xydict)
folddict[file1] = x_foldchange
folddict[file2] = y_foldchange
fold_df = pd.DataFrame(folddict)
interestingdict[file1] = x_interesting
interestingdict[file2] = y_interesting
interesting_df = pd.DataFrame(interestingdict)
sns.set_style("white")
plt.figure(figsize=(4, 4))
plot1 = sns.regplot(x=file1, y=file2, data=xy_df, color=".1", marker='.')
#plot1.set_axis_labels(fnamelist[0] + ' log2(RPKM)', fnamelist[1] + ' log2(RPKM)')
plot1.x = fold_df[file1]
plot1.y = fold_df[file2]
#plot1.plot_joint(plt.scatter, marker='.', c='r', alpha = .4)
plot1.x = interesting_df[file1]
plot1.y = interesting_df[file2]
#plot1.plot_joint(plt.scatter, marker='.', c='b', alpha = .8)
plt.savefig(path_figure + 'RPKM_Cm_media_plot.pdf',
dpi=400,
bbox_inches="tight")
plt.show()
def plot_pausescore(inputs, paths_in, paths_out, settings, settings_plot):
files = inputs['files']
aa_codon = settings_plot['aa_or_codon']
aminoacids = settings_plot['amino_acid']
codons = settings_plot['codon']
ymax_dot = settings_plot['ymax_dot']
ymax_line = settings_plot['ymax_line']
vmax_HM = settings_plot['vmax_HM']
path_figure = paths_out['path_figures']
aa_plots = len(aminoacids)
codon_plots = len(codons)
aa_code, codon_code = ribo_util.get_genetic_code()
for fname in files:
minlength = settings['minlength']
maxlength = settings['maxlength']
plot_upstream = settings[
'plot_upstream'] / 3 * 3 #change window to interval of 3
plot_downstream = settings['plot_downstream'] / 3 * 3
start_trim = settings['start_trim'] / 3 * 3
stop_trim = settings['stop_trim'] / 3 * 3
frameshift = settings['frameshift']
A_site = settings['A_site shift']
minlength_1 = str(minlength) + '_'
maxlength_1 = str(maxlength) + '_'
plot_upstream_1 = str(plot_upstream) + '_'
plot_downstream_1 = str(plot_downstream) + '_'
start_trim_1 = str(start_trim) + '_'
stop_trim_1 = str(stop_trim) + '_'
frameshift_1 = str(frameshift) + '_'
a_site_1 = str(A_site) + '_'
name_settings = minlength_1 + maxlength_1 + plot_upstream_1 + plot_downstream_1
name_settings += start_trim_1 + stop_trim_1 + frameshift_1
path_pausescore = paths_out['path_analysis'] + fname + '/pause_score/'
### For aa_analysis ###
if aa_codon == 'aa':
aa_score = ribo_util.unPickle(path_pausescore + 'aa_scores' +
name_settings)
aa_HM = ribo_util.unPickle(path_pausescore + 'aa_HM_data' +
name_settings)
aa_plot = ribo_util.unPickle(path_pausescore + 'aa_plot_data' +
name_settings)
aa_df = pd.DataFrame(aa_score)
aa_df = aa_df.sort_values(by=['Amino Acid'])
sns.set_style("white")
sns.set_context("talk")
plt.figure(figsize=(8 + 4 * aa_plots, 5))
plt.subplot2grid((2, 2 + aa_plots), (0, 0), rowspan=2, colspan=2)
plot = sns.stripplot(x="Amino Acid",
y="A_site",
data=aa_df,
size=12)
plot = sns.stripplot(x="Amino Acid",
y="P_site",
data=aa_df,
size=6,
color='black')
plot = sns.stripplot(x="Amino Acid",
y="E_site",
data=aa_df,
size=6,
color='grey')
#sns.despine(offset=5, trim = True )
plt.ylim(0, ymax_dot)
plot.axhline(y=1,
xmin=0,
xmax=1,
dashes=[2, 2, 2, 2],
color='grey')
plt.title(fname + ' Amino Acid Pause Scores')
plt.xlabel("Amino Acid")
plt.ylabel("Pause Score")
aa_HM_dict = {}
aa_plot_dict = {}
for aa in aminoacids:
df_HM = pd.DataFrame(aa_HM[aa])
df_plot = pd.DataFrame(aa_plot[aa])
aa_HM_dict[aa] = df_HM
aa_plot_dict[aa] = df_plot
xlim_lower = aa_plot_dict[aa].index[0]
xlim_upper = aa_plot_dict[aa].index[-1]
sns.set_style("white")
sns.set_style("ticks")
plotnum = 0
for aa in aminoacids:
plt.subplot2grid((2, 2 + aa_plots), (0, 2 + plotnum),
rowspan=1,
colspan=1)
plot_2 = plt.plot(aa_plot_dict[aa], sns.xkcd_rgb["dark grey"])
plt.title(aa + ' Average Plot')
plt.ylabel("Pause Score")
plt.ylim(0, ymax_line)
plt.xlim(xlim_lower, xlim_upper)
sns.despine(offset=5)
plt.subplot2grid((2, 2 + aa_plots), (1, 2 + plotnum),
rowspan=1,
colspan=1)
plot_3 = sns.heatmap(aa_HM_dict[aa],
cmap="ocean_r",
vmin=0,
vmax=vmax_HM,
cbar=False,
xticklabels=15,
yticklabels=6)
sns.despine(offset=5)
plotnum += 1
plt.tight_layout()
plt.savefig(path_figure + fname + '/aa_pausescore' +
name_settings + '_aa_pause_scores.png',
dpi=400)
plt.show()
aa_plot_csv = pd.DataFrame(aa_plot)
aa_plot_csv.to_csv(path_pausescore + 'aa_plot_values.csv')
### For codon_analysis ###
if aa_codon == 'codon':
codon_score = ribo_util.unPickle(path_pausescore + 'codon_scores' +
name_settings)
codon_HM = ribo_util.unPickle(path_pausescore + 'codon_HM_data' +
name_settings)
codon_plot = ribo_util.unPickle(path_pausescore +
'codon_plot_data' + name_settings)
aa_list = []
for codon in codon_score['Codon']:
aa = codon_code[codon]
aa_list.append(aa)
codon_score['Amino_Acid'] = aa_list
codon_df = pd.DataFrame(codon_score)
codon_df = codon_df.sort_values(by=['Amino_Acid', 'Codon'])
sns.set_style("white")
sns.set_context("talk")
plt.figure(figsize=(28 + 4 * codon_plots, 5))
plt.subplot2grid((2, 7 + codon_plots), (0, 0),
rowspan=2,
colspan=7)
plot = sns.stripplot(x="Codon", y="A_site", data=codon_df, size=12)
plot = sns.stripplot(x="Codon",
y="P_site",
data=codon_df,
size=6,
color='black')
plot = sns.stripplot(x="Codon",
y="E_site",
data=codon_df,
size=6,
color='grey')
#sns.despine(offset=5, trim = True )
plt.ylim(0, ymax_dot)
plot.axhline(y=1,
xmin=0,
xmax=1,
dashes=[2, 2, 2, 2],
color='grey')
plot.axvline(x=3.5, ymin=0, ymax=ymax_dot, color='grey', lw=.8)
plt.text(1, 4, 'A', fontsize=30) #add text
plot.axvline(x=5.5, ymin=0, ymax=ymax_dot, color='grey', lw=.8)
plt.text(4, 4, 'C', fontsize=30) #add text
plot.axvline(x=7.5, ymin=0, ymax=ymax_dot, color='grey', lw=.8)
plt.text(6, 4, 'D', fontsize=30) #add text
plot.axvline(x=9.5, ymin=0, ymax=ymax_dot, color='grey', lw=.8)
plt.text(8, 4, 'E', fontsize=30) #add text
plot.axvline(x=11.5, ymin=0, ymax=ymax_dot, color='grey', lw=.8)
plt.text(10, 4, 'F', fontsize=30) #add text
plot.axvline(x=15.5, ymin=0, ymax=ymax_dot, color='grey', lw=.8)
plt.text(13, 4, 'G', fontsize=30) #add text
plot.axvline(x=17.5, ymin=0, ymax=ymax_dot, color='grey', lw=.8)
plt.text(16, 4, 'H', fontsize=30) #add text
plot.axvline(x=20.5, ymin=0, ymax=ymax_dot, color='grey', lw=.8)
plt.text(18.8, 4, 'I', fontsize=30) #add text
plot.axvline(x=22.5, ymin=0, ymax=ymax_dot, color='grey', lw=.8)
plt.text(21, 4, 'K', fontsize=30) #add text
plot.axvline(x=28.5, ymin=0, ymax=ymax_dot, color='grey', lw=.8)
plt.text(25, 4, 'L', fontsize=30) #add text
plot.axvline(x=29.5, ymin=0, ymax=ymax_dot, color='grey', lw=.8)
plt.text(28.6, 4, 'M', fontsize=30) #add text
plot.axvline(x=31.5, ymin=0, ymax=ymax_dot, color='grey', lw=.8)
plt.text(30, 4, 'N', fontsize=30) #add text
plot.axvline(x=35.5, ymin=0, ymax=ymax_dot, color='grey', lw=.8)
plt.text(33, 4, 'P', fontsize=30) #add text
plot.axvline(x=37.5, ymin=0, ymax=ymax_dot, color='grey', lw=.8)
plt.text(36, 4, 'Q', fontsize=30) #add text
plot.axvline(x=43.5, ymin=0, ymax=ymax_dot, color='grey', lw=.8)
plt.text(40, 4, 'R', fontsize=30) #add text
plot.axvline(x=49.5, ymin=0, ymax=ymax_dot, color='grey', lw=.8)
plt.text(46, 4, 'S', fontsize=30) #add text
plot.axvline(x=53.5, ymin=0, ymax=ymax_dot, color='grey', lw=.8)
plt.text(51, 4, 'T', fontsize=30) #add text
plot.axvline(x=57.5, ymin=0, ymax=ymax_dot, color='grey', lw=.8)
plt.text(55, 4, 'V', fontsize=30) #add text
plot.axvline(x=60.5, ymin=0, ymax=ymax_dot, color='grey', lw=.8)
plt.text(58.4, 4, 'W', fontsize=30) #add text
plt.text(61.5, 4, '_', fontsize=30) #add text'''
plt.title(fname + ' Codon Pause Scores')
plt.xlabel("Codon")
plt.ylabel("Pause Score")
codon_HM_dict = {}
codon_plot_dict = {}
for codon in codons:
df_HM = pd.DataFrame(codon_HM[codon])
df_plot = pd.DataFrame(codon_plot[codon])
codon_HM_dict[codon] = df_HM
codon_plot_dict[codon] = df_plot
xlim_lower = codon_plot_dict[codon].index[0]
xlim_upper = codon_plot_dict[codon].index[-1]
sns.set_style("white")
sns.set_style("ticks")
plotnum = 0
for codon in codons:
plt.subplot2grid((2, 7 + codon_plots), (0, 7 + plotnum),
rowspan=1,
colspan=1)
plot_2 = plt.plot(codon_plot_dict[codon],
sns.xkcd_rgb["dark grey"])
plt.title(codon + ' Average Plot')
plt.ylabel("Pause Score")
plt.ylim(0, ymax_line)
plt.xlim(xlim_lower, xlim_upper)
sns.despine(offset=5)
plt.subplot2grid((2, 7 + codon_plots), (1, 7 + plotnum),
rowspan=1,
colspan=1)
plot_3 = sns.heatmap(codon_HM_dict[codon],
cmap="ocean_r",
vmin=0,
vmax=vmax_HM,
cbar=False,
xticklabels=15,
yticklabels=6)
sns.despine(offset=5)
plotnum += 1
plt.tight_layout()
plt.savefig(path_figure + fname + '/codon_pausescore' +
name_settings + '.pdf',
dpi=400)
plt.show()
def plot_frame(inputs, paths_in, paths_out, settings, settings_plot):
files = inputs['files']
path_figure = paths_out['path_figures']
minlength = settings['minlength']
maxlength = settings['maxlength']
threshold = settings['threshold']
minlength_1 = str(minlength) + '_'
maxlength_1 = str(maxlength) + '_'
threshold_1 = str(threshold) + '_'
name_settings = minlength_1 + maxlength_1 + threshold_1
for fname in files:
sns.set_style("white")
plt.figure(figsize=(50, 3))
path_analysis = paths_out['path_analysis'] + fname + '/frame/'
frame_alias = ribo_util.unPickle(path_analysis + name_settings +
'frame_alias')
alias_F = []
alias_V = []
alias_A = []
alias_R = []
for alias in frame_alias.keys():
for frame in range(0, 3):
alias_F.append(frame)
alias_V.append(frame_alias[alias][frame])
alias_A.append(alias)
alias_df = pd.DataFrame({
'Frame': alias_F,
'Fraction': alias_V,
'Alias': alias_A
})
frame_length = ribo_util.unPickle(path_analysis + name_settings +
'frame_length')
length_F = []
length_V = []
length_L = []
for length in frame_length.keys():
for frame in range(0, 3):
length_F.append(frame)
length_V.append(frame_length[length][frame])
length_L.append(length)
length_df = pd.DataFrame({
'Frame': length_F,
'Fraction': length_V,
'Length': length_L
})
frame_genome = ribo_util.unPickle(path_analysis + name_settings +
'frame_genome')
x_genome = [0, 1, 2]
plt.subplot(1, 5, 1)
plot1 = sns.barplot(x=x_genome, y=frame_genome, color=".2")
plt.ylabel("Fraction")
plt.xlabel("Frame")
total = float(len(frame_genome))
for p in plot1.patches:
height = p.get_height()
plot1.annotate(
"%.1f" % p.get_height(),
(p.get_x() + p.get_width() / 2., p.get_height() - 10),
ha='center',
va='center',
fontsize=10,
color='.9',
xytext=(0, 20),
textcoords='offset points')
plt.subplot(1, 5, 3)
plot2 = sns.stripplot(x='Length',
y='Fraction',
hue='Frame',
data=length_df,
jitter=False,
size=6,
color="0",
linewidth=0)
plot2.legend_.remove()
#plot2.axhline(y=45, xmin=0, xmax=1, dashes=[2, 1, 2, 1])
frame_0 = length_df.loc[length_df.Frame == 0]
x = frame_0.Length.values
y = frame_0.Fraction.values
plot2.plot(y, sns.xkcd_rgb["pale red"], marker='o')
frame_1 = length_df.loc[length_df.Frame == 1]
x = frame_1.Length.values
y = frame_1.Fraction.values
plot2.plot(y, c='0.3', marker='o')
plt.ylim(0, 100)
plt.title(fname + ' Frame for each Readlength')
plt.subplot(1, 5, 5)
plot3 = sns.boxplot(x='Frame', y='Fraction', data=alias_df, color="c")
plt.title(fname + ' Frame for each Alias')
plot3_1 = sns.stripplot(x='Frame',
y='Fraction',
data=alias_df,
jitter=True,
size=3,
color=".2",
linewidth=0)
plt.savefig(path_figure + fname + '/frame_' + name_settings + '.pdf',
dpi=400)
plt.show()
def plot_avggene(inputs, paths_in, paths_out, settings, settings_plot):
files = inputs['files']
shift = settings_plot['shift']
hmmax = settings_plot['HM_max']
ymax = settings_plot['ymax']
path_figure = paths_out['path_figures']
minlength = settings['minlength']
maxlength = settings['maxlength']
length_in_ORF = settings['length_in_ORF']
length_out_ORF = settings['length_out_ORF']
density_type = settings['density_type']
next_gene = settings['next_gene']
equal_weight = settings['equal_weight']
threshold = settings['threshold']
minlength_1 = str(minlength) + '_'
maxlength_1 = str(maxlength) + '_'
length_in_ORF_1 = str(length_in_ORF) + '_'
length_out_ORF_1 = str(length_out_ORF) + '_'
density_type_1 = density_type + '_'
next_gene_1 = str(next_gene) + '_'
equal_weight_1 = equal_weight + '_'
threshold_1 = str(threshold) + '_'
name_settings = length_in_ORF_1 + length_out_ORF_1 + next_gene_1 + threshold_1
name_settings += density_type_1 + equal_weight_1 + minlength_1 + maxlength_1
for fname in files:
plot_num = 0
sns.set_style("white")
plt.figure(figsize=(20, 5))
path_analysis = paths_out['path_analysis'] + fname + '/avggenes/'
data_start = ribo_util.unPickle(path_analysis + 'avg_start_' +
name_settings + '_all')
data_stop = ribo_util.unPickle(path_analysis + 'avg_stop_' +
name_settings + '_all')
data_startHM = ribo_util.unPickle(path_analysis + 'avg_start_' +
name_settings + '_HM')
data_stopHM = ribo_util.unPickle(path_analysis + 'avg_stop_' +
name_settings + '_HM')
xmax = len(data_start.keys())
data_start = ribo_util.dict_to_df(data_start, 'Position', 'Reads')
data_stop = ribo_util.dict_to_df(data_stop, 'Position', 'Reads')
data_startHM = ribo_util.heatmapdict_to_df(data_startHM, 'Length',
'Position', 'composition')
data_stopHM = ribo_util.heatmapdict_to_df(data_stopHM, 'Length',
'Position', 'composition')
data_startHM = data_startHM.reindex(index=data_startHM.index[::-1])
data_stopHM = data_stopHM.reindex(index=data_stopHM.index[::-1])
data_start.to_csv(path_analysis + 'start_all.csv')
data_stop.to_csv(path_analysis + 'stop_all.csv')
max_start = data_start["Reads"].max()
max_stop = data_stop["Reads"].max()
if ymax == 0:
if max_start > max_stop:
ymax = max_start
else:
ymax = max_stop
for graph in ['Start', 'Stop']:
plot_num += 1
if graph == 'Start':
data = data_start
elif graph == 'Stop':
data = data_stop
plt.subplot(2, 2, plot_num)
plt.plot(
data,
sns.xkcd_rgb["dark grey"],
)
plt.title(fname + ' ' + graph)
plt.ylabel("Reads")
plt.ylim(0, ymax)
plt.xlim(0, xmax)
sns.despine()
for graph in ['startHM', 'stopHM']:
plot_num += 1
if graph == 'startHM':
dataHM = data_startHM
elif graph == 'stopHM':
dataHM = data_stopHM
plt.subplot(2, 2, plot_num)
plot = sns.heatmap(dataHM,
cmap="ocean_r",
vmin=0,
vmax=hmmax,
cbar=False)
plt.setp(plot.get_xticklabels(), visible=False)
plt.setp(plot.get_xticklabels()[0::10], visible=True)
plt.setp(plot.get_yticklabels(), visible=False)
plt.setp(plot.get_yticklabels()[0::4], visible=True)
plt.savefig(path_figure + fname + '/avggene_' + name_settings + '.pdf',
dpi=400)
plt.show()
def plot_pausescore_downstream(inputs, paths_in, paths_out, settings,
settings_plot):
files = inputs['files']
library_id = pd.read_csv(paths_in['files'])
aminoacids = settings_plot['amino_acid']
center_HM = settings_plot['center_HM']
vmax_HM = settings_plot['vmax_HM']
datadict = {}
lib_names = []
plot_data = []
amino_acid = []
position = []
plots = 0
for fname in library_id.Library:
lib = library_id.loc[library_id.Library == fname]
lib_index = lib.Name.index[0]
name = lib.Name.loc[lib_index]
sort = lib.Sort.loc[lib_index]
minlength = settings['minlength']
maxlength = settings['maxlength']
plot_upstream = settings[
'plot_upstream_wave'] / 3 * 3 #change window to interval of 3
plot_downstream = settings['plot_downstream_wave'] / 3 * 3
start_trim = settings['start_trim'] / 3 * 3
stop_trim = settings['stop_trim'] / 3 * 3
frameshift = settings['frameshift']
next_codon = settings['next_codon']
minlength_1 = str(minlength) + '_'
maxlength_1 = str(maxlength) + '_'
plot_upstream_1 = str(plot_upstream) + '_'
plot_downstream_1 = str(plot_downstream) + '_'
start_trim_1 = str(start_trim) + '_'
stop_trim_1 = str(stop_trim) + '_'
frameshift_1 = str(frameshift) + '_'
next_codon_1 = str(next_codon) + '_'
name_settings = plot_upstream_1 + plot_downstream_1 + start_trim_1 + stop_trim_1
name_settings += minlength_1 + maxlength_1 + frameshift_1 + next_codon_1
path_pausescore = paths_out[
'path_analysis'] + fname + '/pause_score/waves/'
aa_plot = ribo_util.unPickle(path_pausescore + name_settings +
'aa_plot_data')
df_plot = pd.DataFrame(aa_plot)
for aa in aminoacids:
data = df_plot[aa].tolist()
data_length = len(data)
data_index = 0
for index in range(0, data_length / 3):
codon_value = data[data_index - 1:data_index + 2]
codon_value = sum(codon_value) / 3
lib_names.append(sort)
plot_data.append(codon_value)
amino_acid.append(aa)
position.append(index)
data_index += 3
plots += 1
datadict['Sort'] = lib_names
datadict['Data'] = plot_data
datadict['AA'] = amino_acid
datadict['Position'] = position
data_df = pd.DataFrame(datadict)
for aa in aminoacids:
aa_data = data_df.loc[data_df.AA == aa]
aa_data = aa_data.drop('AA', axis=1)
aa_data = aa_data.pivot('Sort', 'Position', 'Data')
sns.set_style("white")
sns.set_context("talk")
plt.figure(figsize=(20, .8 + .4 * plots))
vmax = vmax_HM
center = center_HM
vmin = center_HM - (vmax_HM - center)
plot_A = sns.heatmap(
aa_data,
cmap="RdBu_r",
vmax=vmax,
center=center,
vmin=vmin,
cbar=True,
)
plt.title(aa)
outpath = paths_out['path_figures']
plt.savefig(outpath + '/Comparison/' + aa + 'pause_wave.png', dpi=400)
plt.show()
def plot_pausescore_APE_heatmap(inputs, paths_in, paths_out, settings,
settings_plot):
files = inputs['files']
library_id = pd.read_csv(paths_in['files'])
center_HM = settings_plot['center_HM']
vmax_HM = settings_plot['vmax_HM']
namelist = []
library = []
Amino_acid = []
A_score = []
P_score = []
E_score = []
A_dict = {}
P_dict = {}
E_dict = {}
all_dict = {}
plots = 0
for fname in library_id.Library:
lib = library_id.loc[library_id.Library == fname]
lib_index = lib.Name.index[0]
name = lib.Name.loc[lib_index]
sort = lib.Sort.loc[lib_index]
minlength = settings['minlength']
maxlength = settings['maxlength']
plot_upstream = settings[
'plot_upstream'] / 3 * 3 #change window to interval of 3
plot_downstream = settings['plot_downstream'] / 3 * 3
start_trim = settings['start_trim'] / 3 * 3
stop_trim = settings['stop_trim'] / 3 * 3
frameshift = settings['frameshift']
minlength_1 = str(minlength) + '_'
maxlength_1 = str(maxlength) + '_'
plot_upstream_1 = str(plot_upstream) + '_'
plot_downstream_1 = str(plot_downstream) + '_'
start_trim_1 = str(start_trim) + '_'
stop_trim_1 = str(stop_trim) + '_'
frameshift_1 = str(frameshift) + '_'
name_settings = plot_upstream_1 + plot_downstream_1 + start_trim_1 + stop_trim_1 + minlength_1 + maxlength_1 + frameshift_1
path_pausescore = paths_out['path_analysis'] + fname + '/pause_score/'
aa_score = ribo_util.unPickle(path_pausescore + name_settings +
'aa_scores')
aa_HM = ribo_util.unPickle(path_pausescore + name_settings +
'aa_HM_data')
aa_plot = ribo_util.unPickle(path_pausescore + name_settings +
'aa_plot_data')
amino_acids = aa_score['Amino Acid']
a_site = aa_score['A_site']
p_site = aa_score['P_site']
e_site = aa_score['E_site']
plots += 1
for aa, ascore, pscore, escore in itertools.izip(
amino_acids, a_site, p_site, e_site):
if aa == '_':
continue
namelist.append(fname)
library.append(sort)
Amino_acid.append(aa)
A_score.append(ascore)
P_score.append(pscore)
E_score.append(escore)
A_dict['Library'] = library
A_dict['Amino Acid'] = Amino_acid
A_dict['A site'] = A_score
P_dict['Library'] = library
P_dict['Amino Acid'] = Amino_acid
P_dict['P site'] = P_score
E_dict['Library'] = library
E_dict['Amino Acid'] = Amino_acid
E_dict['E site'] = E_score
all_dict['Library'] = namelist
all_dict['Amino Acid'] = Amino_acid
all_dict['A site'] = A_score
all_dict['P site'] = P_score
all_dict['E site'] = E_score
A_df = pd.DataFrame(A_dict)
P_df = pd.DataFrame(P_dict)
E_df = pd.DataFrame(E_dict)
all_df = pd.DataFrame(all_dict)
A_df = A_df.pivot('Library', 'Amino Acid', 'A site')
P_df = P_df.pivot('Library', 'Amino Acid', 'P site')
E_df = E_df.pivot('Library', 'Amino Acid', 'E site')
all_df = all_df[['Library', 'Amino Acid', 'A site', 'P site', 'E site']]
sns.set_style("white")
sns.set_context("talk")
plt.figure(figsize=(20, .8 + .4 * plots))
plt.subplot(1, 3, 1)
plot_A = sns.heatmap(
A_df,
cmap="RdBu_r",
vmax=vmax_HM,
center=center_HM,
cbar=False,
)
plt.title('A-Site Pause Scores')
plt.subplot(1, 3, 2)
plot_P = sns.heatmap(
P_df,
cmap="RdBu_r",
vmax=vmax_HM,
center=center_HM,
cbar=False,
)
plt.title('P-Site Pause Scores')
plt.subplot(1, 3, 3)
plot_E = sns.heatmap(
E_df,
cmap="RdBu_r",
vmax=vmax_HM,
center=center_HM,
cbar=True,
)
plt.title('E-Site Pause Scores')
plt.tight_layout()
outpath = paths_out['path_figures']
plt.savefig(outpath + '/Comparison/' + 'pause_scores.png', dpi=400)
plt.show()
def plot_avggene_end(inputs, paths_in, paths_out, settings):
files = inputs['files']
shift = settings['shift']
hmmax = settings['HM_max']
for fname in files:
plot_num = 0
plt.figure(figsize=(20, 5))
path_analysis = paths_out['path_analysis'] + fname + '/'
data_start = ribo_util.unPickle(path_analysis + 'avg_start_all_end')
data_stop = ribo_util.unPickle(path_analysis + 'avg_stop_all_end')
data_startHM = ribo_util.unPickle(path_analysis + 'avg_start_HM_end')
data_stopHM = ribo_util.unPickle(path_analysis + 'avg_stop_HM_end')
xmax = len(data_start.keys())
data_start = ribo_util.dict_to_df(data_start, 'Position', 'Reads')
data_stop = ribo_util.dict_to_df(data_stop, 'Position', 'Reads')
data_startHM = ribo_util.heatmapdict_to_df(data_startHM, 'Length',
'Position', 'composition')
data_stopHM = ribo_util.heatmapdict_to_df(data_stopHM, 'Length',
'Position', 'composition')
data_start.to_csv(path_analysis + 'start_all.csv')
data_stop.to_csv(path_analysis + 'stop_all.csv')
max_start = data_start["Reads"].max()
max_stop = data_stop["Reads"].max()
if max_start > max_stop:
ymax = max_start
else:
ymax = max_stop
for graph in ['Start', 'Stop']:
plot_num += 1
if graph == 'Start':
data = data_start
elif graph == 'Stop':
data = data_stop
plt.subplot(2, 2, plot_num)
plt.plot(
data,
sns.xkcd_rgb["dark grey"],
)
plt.title(fname + ' ' + graph)
plt.ylabel("Reads")
plt.ylim(0, ymax)
plt.xlim(0, xmax)
sns.despine()
for graph in ['startHM', 'stopHM']:
plot_num += 1
if graph == 'startHM':
dataHM = data_startHM
elif graph == 'stopHM':
dataHM = data_stopHM
plt.subplot(2, 2, plot_num)
plot = sns.heatmap(dataHM,
cmap="ocean_r",
vmin=0,
vmax=hmmax,
cbar=False)
plt.setp(plot.get_xticklabels(), visible=False)
plt.setp(plot.get_xticklabels()[0::10], visible=True)
plt.setp(plot.get_yticklabels(), visible=False)
plt.setp(plot.get_yticklabels()[0::4], visible=True)
plt.show()
def plot_TEvsSD(fnamelist, paths_in, paths_out, settings):
'''Takes 2 files and compares translation efficiency
to shine dalgarno affinity'''
'''Load Settings'''
foldchange = settings['foldchange']
highlight_genes = settings['interesting']
data_transform = settings[
'data_transform'] # has to be 'None', 'log10', or 'log2'
'''Load Data'''
#genelist file from analysis
path_analysis1 = paths_out['path_analysis'] + fnamelist[0] + '/'
path_analysis2 = paths_out['path_analysis'] + fnamelist[1] + '/'
genelist_f1 = ribo_util.unPickle(path_analysis1 + 'genelist')
genelist_f2 = ribo_util.unPickle(path_analysis2 + 'genelist')
#define alias and RPKM as arrays
df1 = pd.DataFrame(genelist_f1)
alias = df1.Alias
RPKM_f1 = df1.RPKM
df2 = pd.DataFrame(genelist_f2)
RPKM_f2 = df2.RPKM
'''data arrays to plot'''
x = []
y = []
x_foldchange = []
y_foldchange = []
x_interesting = []
y_interesting = []
'''iterate through gene data for manipulation (log transform and fold change calc)'''
for alias, xval, yval in itertools.izip(alias.values, RPKM_f1.values,
RPKM_f2.values):
# remove genes with RPKM = 0
if xval == 0 or yval == 0:
continue
# calculate foldchange prior to data transformation
foldchange = abs(yval - xval)
# transform data
if data_transform == 'None':
xval = xval
yval = yval
if data_transform == 'log2':
xval = math.log2(xval)
yval = math.log2(yval)
if data_transform == 'log10':
xval = math.log10(xval)
yval = math.log10(yval)
# append data to lists
x.append(xval)
y.append(yval)
if foldchange > settings['foldchange']:
x_foldchange.append(xval)
y_foldchange.append(yval)
if alias in settings['interesting']:
x_interesting.append(xval)
y_interesting.append(yval)
plot1 = sns.jointplot(x=x, y=y, size=6, color=".4", marker='.', alpha=.5)
plot1.set_axis_labels(fnamelist[0] + ' log2(RPKM)',
fnamelist[1] + ' log2(RPKM)')
plot1.x = x_foldchange
plot1.y = y_foldchange
plot1.plot_joint(plt.scatter, marker='.', c='r', alpha=.4)
plot1.x = x_interesting
plot1.y = y_interesting
plot1.plot_joint(plt.scatter, marker='.', c='b', alpha=.8)
plt.show()