def plot_distributions(meme_gerp_scores, null_gerp_scores, mylabels, mytitle):
for gerp_scores, mylabel in zip([meme_gerp_scores, null_gerp_scores],
mylabels):
plot_functions.plot_density(gerp_scores, mytitle, mylabel)
plt.legend()
plt.show()
def plot_distributions(meme_gerp_scores, null_gerp_scores, mylabels, mytitle):
for gerp_scores, mylabel in zip([meme_gerp_scores,
null_gerp_scores],
mylabels):
plot_functions.plot_density(gerp_scores, mytitle, mylabel)
plt.legend()
plt.show()
def main():
usage = 'usage: %prog [opt] directory1 directory2'\
'\nTwo arguments must be specified in command line:\n'\
'1) Directory of interest containing .summary files\n'\
'2) Directory to which to compare containing .summary files (control)\n'
parser = OptionParser(usage=usage)
(_, args) = parser.parse_args()
if len(args) != 2:
print 'Requires 2 arguments to be specified in command line'
print usage
sys.exit()
# parse args
dir0 = args[0]
dir1 = args[1]
# get all files containing *.summary in each directory. Those are
# anchor input files.
ext = '.summary'
anchor_files0 = \
[os.path.join(dir0, f) for f in os.listdir(dir0) if f.endswith(ext)]
anchor_files1 = \
[os.path.join(dir1, f) for f in os.listdir(dir1) if f.endswith(ext)]
# Read each file, retrieving the lengths of each amino acid sequence
aa_lengths0 = []
aa_lengths1 = []
for file0 in anchor_files0:
aa_lengths0 += get_aa_length_from_anchor_file(file0)
for file1 in anchor_files1:
aa_lengths1 += get_aa_length_from_anchor_file(file1)
for id, aa in zip(['xeno', 'control'], [aa_lengths0, aa_lengths1]):
print 'Mean for %s' % id
print sum(aa) / float(len(aa))
print 'Median for %s' % id
print sorted(aa)[len(aa) // 2]
print '[min,max] for %s' % id
print '[%s,%s]' % (min(aa), max(aa))
plot_functions.plot_density(
[aa_lengths0, aa_lengths1],
mytitle='Density plot of exon lengths',
xlabel='Nucleotide length',
ylabel='Density',
labels_lists=['Cassette Exons', 'Constitutive Exons'],
smoothness=0.1,
legend_pos=1,
ymin=0,
ymax=0.025,
xmin=0,
xmax=200)
def main():
usage = 'usage: %prog [opt] directory1 directory2'\
'\nTwo arguments must be specified in command line:\n'\
'1) Directory of interest containing .summary files\n'\
'2) Directory to which to compare containing .summary files (control)\n'
parser = OptionParser(usage=usage)
(_, args) = parser.parse_args()
if len(args) != 2:
print 'Requires 2 arguments to be specified in command line'
print usage
sys.exit()
# parse args
dir0 = args[0]
dir1 = args[1]
# get all files containing *.summary in each directory. Those are
# anchor input files.
ext = '.summary'
anchor_files0 = \
[os.path.join(dir0, f) for f in os.listdir(dir0) if f.endswith(ext)]
anchor_files1 = \
[os.path.join(dir1, f) for f in os.listdir(dir1) if f.endswith(ext)]
# Read each file, retrieving the lengths of each amino acid sequence
aa_lengths0 = []
aa_lengths1 = []
for file0 in anchor_files0:
aa_lengths0 += get_aa_length_from_anchor_file(file0)
for file1 in anchor_files1:
aa_lengths1 += get_aa_length_from_anchor_file(file1)
for id, aa in zip(['xeno', 'control'], [aa_lengths0, aa_lengths1]):
print 'Mean for %s' %id
print sum(aa)/float(len(aa))
print 'Median for %s' %id
print sorted(aa)[len(aa)//2]
print '[min,max] for %s' %id
print '[%s,%s]' %(min(aa), max(aa))
plot_functions.plot_density([aa_lengths0, aa_lengths1],
mytitle='Density plot of exon lengths',
xlabel='Nucleotide length',
ylabel='Density',
labels_lists=['Cassette Exons',
'Constitutive Exons'],
smoothness=0.1,
legend_pos=1,
ymin=0, ymax=0.025,
xmin=0, xmax=200)
def main():
usage = 'usage: %prog meme_gerp_summary_pkl\n'\
'Requires one argument:\n'\
'1) pkl file from summarize_meme_results'
parser = OptionParser(usage=usage)
parser.add_option('-r', '--plot_raw_locations', dest='raw_locations',
default=False,
help='Boolean value. True=horizontal '\
'line segment plot. False=density plot')
(options, args) = parser.parse_args()
if len(args) != 1:
print 'Requires 1 argument to be specified in commandline'
print usage
sys.exit()
pklpath = args[0]
if options.raw_locations in ['True', 'TRUE', True]:
raw_locations = True
print 'Plotting raw locations...'
elif options.raw_locations in ['False', 'FALSE', False]:
raw_locations = False
print 'Plotting density plot...'
else:
print '--plot_raw_locations option must be '\
'True or False. %s found.' %options.raw_locations
sys.exit()
# get dics from pkl
meme_dic = get_dic_from_pklpath(pklpath)
print meme_dic
event_count = 0 # used as y-axis locater...
# init offsetters
offset_length = 100
offsets = [0, 110, 220, 330, 440]
'''
Set Motif 1 to #CC6666, Motif 2 to #33CCCC Motif 3 to "green"
The colors you want will depend on the discovered meme motif number.
'''
plot_settings_dic = {'intron_1_5p': {'offset': offsets[0],
'color': ['#CC6666', '#CC6666', '#CC6666']},
'intron_1_3p': {'offset': offsets[1],
'color': ['green', 'black', 'yellow']},
'intron_2_5p': {'offset': offsets[2],
'color': ['red']},
'intron_2_3p': {'offset': offsets[3],
'color': ['#33CCCC', 'black']}}
# collect plot information: start, end, color, y position
# into a plot dic.
plot_dic = {'start': [],
'end': [],
'color': [],
'ypos': [],
'motif_number': []}
for event in meme_dic:
for region in meme_dic[event]:
start = meme_dic[event][region]['motif_relative_start'][0]
end = meme_dic[event][region]['motif_relative_end'][0]
motif_number = meme_dic[event][region]['motif_number'][0]
# offset start and end depending on region
start += plot_settings_dic[region]['offset']
end += plot_settings_dic[region]['offset']
ypos = event_count/10.0
try:
# subtract motif number by 1 to get 0-based numbering
color = plot_settings_dic[region]['color'][motif_number-1]
except IndexError:
print region, motif_number
print 'Ran out of colors, using yellow as default.'
color = 'yellow'
for key, value in \
zip(['start', 'end', 'color', 'ypos', 'motif_number'],
[start, end, color, ypos, '%s:Motif %s'%(region, motif_number)]):
plot_dic[key].append(value)
event_count += 1
if raw_locations is False:
# begin: get lists of starts, colors, labels for density plot
density_plot_dic = {}
for motif_number, color, start in \
zip(plot_dic['motif_number'], plot_dic['color'], plot_dic['start']):
if motif_number not in density_plot_dic:
density_plot_dic[motif_number] = {'densitystarts': []}
density_plot_dic[motif_number]['densitycolors'] = color
density_plot_dic[motif_number]['densitystarts'].append(start)
starts_list = []
labels_list = []
colors_list = []
for motif_number in density_plot_dic:
labels_list.append(motif_number)
colors_list.append(density_plot_dic[motif_number]['densitycolors'])
starts_list.append(density_plot_dic[motif_number]['densitystarts'])
# add number of sites in labels
labels_with_nsites = []
motif_labels = ['Motif %s' %n for n in range(1, len(starts_list) + 1)]
for labellist, startlist in zip(motif_labels, starts_list):
n_sites = len(startlist)
labels_with_nsites.append('%s (%s sites)' %(labellist, n_sites))
for startlist in starts_list:
print 'Number of guys: %s' %len(startlist)
print 'Min/Max: %s/%s' %(min(startlist), max(startlist))
# end: get lists of starts, colors, labels for density plot
# begin: init figure
fig = plt.figure()
ax = fig.add_subplot(111)
# end: init figure
# add rectangles representing exons
rect_height=0.002
rect_length=10
rectstarts = [offsets[0], offsets[2], offsets[4]]
rcolors = ['cyan', 'yellow', 'cyan']
for start, color in zip(rectstarts, rcolors):
patch = add_rectangles(start, height=rect_height, length=rect_length, color=color)
ax.add_patch(patch)
# draw intron lines connecting exons
istarts = [offsets[0], offsets[1], offsets[2], offsets[3]]
iends = [offsets[1] - rect_length, offsets[2] - rect_length, offsets[3] - rect_length, offsets[4] - rect_length]
for start, end in zip(istarts, iends):
plt.hlines(y=-rect_height/2., xmin=start, xmax=end,
color='black', linewidths=1.5)
# draw vertical lines representing break in intron
breakstarts = [iends[0], istarts[1], iends[2], istarts[3]]
for bstart in breakstarts:
plt.vlines(bstart, ymin=-rect_height, ymax=0,
color='black', linewidths=1)
plot_functions.plot_density(values_lists=starts_list,
labels_lists=labels_with_nsites,
colors_list=colors_list,
mytitle='Intronic distribution of MEME motifs',
xlabel='Genic region',
ylabel='Density',
xmin=-20,
xmax=450,
smoothness=0.1,
legend_pos=2,
ymin=-0.01,
ymax=0.075,
showplot=False)
# dont show xaxis
plt.setp(ax.get_xticklabels(), visible=False)
plt.show()
else:
plot_functions.plot_hline_segments(starts=plot_dic['start'],
stops=plot_dic['end'],
ypos=plot_dic['ypos'],
colors=plot_dic['color'],
labels=plot_dic['motif_number'])
def main():
usage = 'usage: %prog meme_gerp_summary_pkl\n'\
'Requires one argument:\n'\
'1) pkl file from summarize_meme_results'
parser = OptionParser(usage=usage)
parser.add_option('-r', '--plot_raw_locations', dest='raw_locations',
default=False,
help='Boolean value. True=horizontal '\
'line segment plot. False=density plot')
(options, args) = parser.parse_args()
if len(args) != 1:
print 'Requires 1 argument to be specified in commandline'
print usage
sys.exit()
pklpath = args[0]
if options.raw_locations in ['True', 'TRUE', True]:
raw_locations = True
print 'Plotting raw locations...'
elif options.raw_locations in ['False', 'FALSE', False]:
raw_locations = False
print 'Plotting density plot...'
else:
print '--plot_raw_locations option must be '\
'True or False. %s found.' %options.raw_locations
sys.exit()
# get dics from pkl
meme_dic = get_dic_from_pklpath(pklpath)
print meme_dic
event_count = 0 # used as y-axis locater...
# init offsetters
offset_length = 100
offsets = [0, 110, 220, 330, 440]
'''
Set Motif 1 to #CC6666, Motif 2 to #33CCCC Motif 3 to "green"
The colors you want will depend on the discovered meme motif number.
'''
plot_settings_dic = {
'intron_1_5p': {
'offset': offsets[0],
'color': ['#CC6666', '#CC6666', '#CC6666']
},
'intron_1_3p': {
'offset': offsets[1],
'color': ['green', 'black', 'yellow']
},
'intron_2_5p': {
'offset': offsets[2],
'color': ['red']
},
'intron_2_3p': {
'offset': offsets[3],
'color': ['#33CCCC', 'black']
}
}
# collect plot information: start, end, color, y position
# into a plot dic.
plot_dic = {
'start': [],
'end': [],
'color': [],
'ypos': [],
'motif_number': []
}
for event in meme_dic:
for region in meme_dic[event]:
start = meme_dic[event][region]['motif_relative_start'][0]
end = meme_dic[event][region]['motif_relative_end'][0]
motif_number = meme_dic[event][region]['motif_number'][0]
# offset start and end depending on region
start += plot_settings_dic[region]['offset']
end += plot_settings_dic[region]['offset']
ypos = event_count / 10.0
try:
# subtract motif number by 1 to get 0-based numbering
color = plot_settings_dic[region]['color'][motif_number - 1]
except IndexError:
print region, motif_number
print 'Ran out of colors, using yellow as default.'
color = 'yellow'
for key, value in \
zip(['start', 'end', 'color', 'ypos', 'motif_number'],
[start, end, color, ypos, '%s:Motif %s'%(region, motif_number)]):
plot_dic[key].append(value)
event_count += 1
if raw_locations is False:
# begin: get lists of starts, colors, labels for density plot
density_plot_dic = {}
for motif_number, color, start in \
zip(plot_dic['motif_number'], plot_dic['color'], plot_dic['start']):
if motif_number not in density_plot_dic:
density_plot_dic[motif_number] = {'densitystarts': []}
density_plot_dic[motif_number]['densitycolors'] = color
density_plot_dic[motif_number]['densitystarts'].append(start)
starts_list = []
labels_list = []
colors_list = []
for motif_number in density_plot_dic:
labels_list.append(motif_number)
colors_list.append(density_plot_dic[motif_number]['densitycolors'])
starts_list.append(density_plot_dic[motif_number]['densitystarts'])
# add number of sites in labels
labels_with_nsites = []
motif_labels = ['Motif %s' % n for n in range(1, len(starts_list) + 1)]
for labellist, startlist in zip(motif_labels, starts_list):
n_sites = len(startlist)
labels_with_nsites.append('%s (%s sites)' % (labellist, n_sites))
for startlist in starts_list:
print 'Number of guys: %s' % len(startlist)
print 'Min/Max: %s/%s' % (min(startlist), max(startlist))
# end: get lists of starts, colors, labels for density plot
# begin: init figure
fig = plt.figure()
ax = fig.add_subplot(111)
# end: init figure
# add rectangles representing exons
rect_height = 0.002
rect_length = 10
rectstarts = [offsets[0], offsets[2], offsets[4]]
rcolors = ['cyan', 'yellow', 'cyan']
for start, color in zip(rectstarts, rcolors):
patch = add_rectangles(start,
height=rect_height,
length=rect_length,
color=color)
ax.add_patch(patch)
# draw intron lines connecting exons
istarts = [offsets[0], offsets[1], offsets[2], offsets[3]]
iends = [
offsets[1] - rect_length, offsets[2] - rect_length,
offsets[3] - rect_length, offsets[4] - rect_length
]
for start, end in zip(istarts, iends):
plt.hlines(y=-rect_height / 2.,
xmin=start,
xmax=end,
color='black',
linewidths=1.5)
# draw vertical lines representing break in intron
breakstarts = [iends[0], istarts[1], iends[2], istarts[3]]
for bstart in breakstarts:
plt.vlines(bstart,
ymin=-rect_height,
ymax=0,
color='black',
linewidths=1)
plot_functions.plot_density(
values_lists=starts_list,
labels_lists=labels_with_nsites,
colors_list=colors_list,
mytitle='Intronic distribution of MEME motifs',
xlabel='Genic region',
ylabel='Density',
xmin=-20,
xmax=450,
smoothness=0.1,
legend_pos=2,
ymin=-0.01,
ymax=0.075,
showplot=False)
# dont show xaxis
plt.setp(ax.get_xticklabels(), visible=False)
plt.show()
else:
plot_functions.plot_hline_segments(starts=plot_dic['start'],
stops=plot_dic['end'],
ypos=plot_dic['ypos'],
colors=plot_dic['color'],
labels=plot_dic['motif_number'])
def main():
usage = 'usage: %prog meme_gerp_genename_filepath output_filepath\n'\
'Requires two input arguments:\n'\
'1) textfile output from '\
'summarize_meme_results_with_gerp_scores\n'\
'2) inclusion fasta file\n'\
'3) exclusion fasta file\n'\
'4) meme dir containing meme results'
parser = OptionParser(usage=usage)
(_, args) = parser.parse_args()
if len(args) < 5:
print 'Four arguments need to be specified in command line.\n'
print usage
sys.exit()
meme_summarypath = args[0]
incl_fasta = args[1]
excl_fasta = args[2]
meme_dir = args[3]
# define column name suffix (string after colon in colname)
gerp_str = 'avg_rs_score'
motif_numb_str = 'motif_number'
# miso event has no col name suffix, this is entire colname
miso_colname = 'miso_event'
# define rel path to tomtom files from meme dir
rel_path = os.path.join('rbp_matches', 'candidate_rbps.txt')
# define plot title
mytitle = 'GERP Score Comparison: Hits vs Non-Hits'
# get dictionary containing inclusion and exclusion for miso event
incl_excl_dic = miso_events.get_inclusion_exclusion(incl_file=incl_fasta,
excl_file=excl_fasta)
tomtom_dic = miso_events.get_tomtom_hits(meme_dir, rel_path)
region_gerp_scores = {} # gerp scores, indexed by region.
with open(meme_summarypath, 'rb') as readfile:
myreader = csv.reader(readfile, delimiter='\t')
header = myreader.next()
regions = get_regions(header)
# init output dic with empty lists
for region in regions:
region_gerp_scores[region] = {}
for row in myreader:
# get gerp score in each region.
# beware of empty values.
for region in regions:
subdic = region_gerp_scores[region]
gerp_colname = ':'.join([region, gerp_str])
motif_numb_colname = ':'.join([region, motif_numb_str])
gerp_score = row[header.index(gerp_colname)]
motif_numb = row[header.index(motif_numb_colname)]
miso_event = row[header.index(miso_colname)]
incl_or_excl = incl_excl_dic[miso_event]
motif_id = ' '.join(['Motif', motif_numb, incl_or_excl])
if gerp_score is not '':
if motif_id not in subdic:
subdic[motif_id] = []
subdic[motif_id].append(float(gerp_score))
avg_scores_in_tomtom = []
avg_scores_not_in_tomtom = []
# Plot histogram of average scores.
for region in region_gerp_scores:
for motif_id in region_gerp_scores[region]:
tomtom_key = create_tomtom_key(motif_id, region)
if tomtom_key in tomtom_dic:
avg_scores_in_tomtom += region_gerp_scores[region][motif_id]
else:
avg_scores_not_in_tomtom += region_gerp_scores[region][
motif_id]
conserved_counts_in_tomtom = 0
conserved_counts_not_in_tomtom = 0
for s in avg_scores_in_tomtom:
if s >= 2:
conserved_counts_in_tomtom += 1
for s in avg_scores_not_in_tomtom:
if s >= 2:
conserved_counts_not_in_tomtom += 1
for avg_scores, mylabel in zip(
[avg_scores_in_tomtom, avg_scores_not_in_tomtom],
['Motif with matching RBPs', 'Motif without matching RBPs']):
plot_functions.plot_density(avg_scores,
mytitle=mytitle,
mylabel=mylabel)
plt.legend()
plt.show()
def main():
usage = 'usage: %prog meme_gerp_genename_filepath output_filepath\n'\
'Requires two input arguments:\n'\
'1) pkl file from summarize_meme_results: non-null\n'\
'2) pkl file from summarize_meme_results: null-mode\n'
parser = OptionParser(usage=usage)
parser.add_option('-t', '--threshold', dest='score_threshold',
default=2.0,
help='Float, threshold for what one considers conserved.')
parser.add_option('-y', '--ymax', dest='ymax',
type='float',
default=0.03,
help='Y max for density plot')
(options, args) = parser.parse_args()
if len(args) < 2:
print 'Two arguments need to be specified in command line.\n'
print usage
sys.exit()
non_null_pklpath = args[0]
null_pklpath = args[1]
# parse ops
score_threshold = float(options.score_threshold)
# get dics from pkl
non_null_dic = get_dic_from_pklpath(non_null_pklpath)
null_dic = get_dic_from_pklpath(null_pklpath)
non_null_gerp_scores = get_gerp_scores(non_null_dic, gerpkey='avg_rs_score')
null_gerp_scores = get_gerp_scores(null_dic, gerpkey='avg_rs_score')
plot_functions.plot_density([non_null_gerp_scores, null_gerp_scores],
mytitle='Density plot of conservation scores',
labels_lists=['MEME motifs', 'Controls'],
xlabel='GERP conservation score',
ylabel='Density',
xmin=-4, xmax=4,
ymax=options.ymax,
smoothness=0.15,
drawvline=score_threshold)
# find how many conserved regions are in each.
n_conserved_in_meme = \
gerp_utilities.conserved_regions(non_null_gerp_scores, fraction=False, threshold=score_threshold)
n_conserved_in_null = \
gerp_utilities.conserved_regions(null_gerp_scores, fraction=False, threshold=score_threshold)
n_total_in_meme = len(non_null_gerp_scores)
n_total_in_null = len(null_gerp_scores)
n_not_conserved_in_meme = n_total_in_meme - n_conserved_in_meme
n_not_conserved_in_null = n_total_in_null - n_conserved_in_null
print 'Threshold: %s' %score_threshold
print 'Number of conserved elements: %s' %n_conserved_in_meme
print 'Number of conserved elements found in control: %s' %n_conserved_in_null
# Perform fisher's exact test
oddsratio, pvalue = fisher_exact([[n_conserved_in_meme,
n_conserved_in_null],
[n_not_conserved_in_meme,
n_not_conserved_in_null]])
print 'Fishers Exact Test, Oddsratio: %s. Pvalue: %s' %(oddsratio, pvalue)
# plot distributions
mylabels = ['Meme motifs', 'Control region']
mytitle = 'Fraction of elements conserved compared to control region'
# Plot bargraphs
frac_conserved_meme = float(n_conserved_in_meme) / n_total_in_meme
frac_conserved_null = float(n_conserved_in_null) / n_total_in_null
myvals = [frac_conserved_meme, frac_conserved_null]
plot_functions.plot_barplot(myvals, mytitle, mylabels,
ylabel='Fraction of elements conserved',
mytext1="%i/%i" \
%(n_conserved_in_meme,
n_total_in_meme),
mytext2='%i/%i' %(n_conserved_in_null,
n_total_in_null),
mytext3="*Fisher's Exact Test\nP-value=%.2e" %pvalue,
ymin=0,
ymax=1,
width=0.5)
plt.show()
def main():
usage = 'usage: %prog meme_gerp_genename_filepath output_filepath\n'\
'Requires two input arguments:\n'\
'1) textfile output from '\
'summarize_meme_results_with_gerp_scores\n'\
'2) inclusion fasta file\n'\
'3) exclusion fasta file\n'\
'4) meme dir containing meme results'
parser = OptionParser(usage=usage)
(_, args) = parser.parse_args()
if len(args) < 5:
print 'Four arguments need to be specified in command line.\n'
print usage
sys.exit()
meme_summarypath = args[0]
incl_fasta = args[1]
excl_fasta = args[2]
meme_dir = args[3]
# define column name suffix (string after colon in colname)
gerp_str = 'avg_rs_score'
motif_numb_str = 'motif_number'
# miso event has no col name suffix, this is entire colname
miso_colname = 'miso_event'
# define rel path to tomtom files from meme dir
rel_path = os.path.join('rbp_matches', 'candidate_rbps.txt')
# define plot title
mytitle = 'GERP Score Comparison: Hits vs Non-Hits'
# get dictionary containing inclusion and exclusion for miso event
incl_excl_dic = miso_events.get_inclusion_exclusion(incl_file=incl_fasta,
excl_file=excl_fasta)
tomtom_dic = miso_events.get_tomtom_hits(meme_dir, rel_path)
region_gerp_scores = {} # gerp scores, indexed by region.
with open(meme_summarypath, 'rb') as readfile:
myreader = csv.reader(readfile, delimiter='\t')
header = myreader.next()
regions = get_regions(header)
# init output dic with empty lists
for region in regions:
region_gerp_scores[region] = {}
for row in myreader:
# get gerp score in each region.
# beware of empty values.
for region in regions:
subdic = region_gerp_scores[region]
gerp_colname = ':'.join([region, gerp_str])
motif_numb_colname = ':'.join([region, motif_numb_str])
gerp_score = row[header.index(gerp_colname)]
motif_numb = row[header.index(motif_numb_colname)]
miso_event = row[header.index(miso_colname)]
incl_or_excl = incl_excl_dic[miso_event]
motif_id = ' '.join(['Motif', motif_numb, incl_or_excl])
if gerp_score is not '':
if motif_id not in subdic:
subdic[motif_id] = []
subdic[motif_id].append(float(gerp_score))
avg_scores_in_tomtom = []
avg_scores_not_in_tomtom = []
# Plot histogram of average scores.
for region in region_gerp_scores:
for motif_id in region_gerp_scores[region]:
tomtom_key = create_tomtom_key(motif_id, region)
if tomtom_key in tomtom_dic:
avg_scores_in_tomtom += region_gerp_scores[region][motif_id]
else:
avg_scores_not_in_tomtom += region_gerp_scores[region][motif_id]
conserved_counts_in_tomtom = 0
conserved_counts_not_in_tomtom = 0
for s in avg_scores_in_tomtom:
if s >= 2:
conserved_counts_in_tomtom += 1
for s in avg_scores_not_in_tomtom:
if s >= 2:
conserved_counts_not_in_tomtom += 1
for avg_scores, mylabel in zip([avg_scores_in_tomtom,
avg_scores_not_in_tomtom],
['Motif with matching RBPs',
'Motif without matching RBPs']):
plot_functions.plot_density(avg_scores,
mytitle=mytitle,
mylabel=mylabel)
plt.legend()
plt.show()
def main():
usage = 'usage: %prog meme_gerp_genename_filepath output_filepath\n'\
'Requires two input arguments:\n'\
'1) pkl file from summarize_meme_results: non-null\n'\
'2) pkl file from summarize_meme_results: null-mode\n'
parser = OptionParser(usage=usage)
parser.add_option(
'-t',
'--threshold',
dest='score_threshold',
default=2.0,
help='Float, threshold for what one considers conserved.')
parser.add_option('-y',
'--ymax',
dest='ymax',
type='float',
default=0.03,
help='Y max for density plot')
(options, args) = parser.parse_args()
if len(args) < 2:
print 'Two arguments need to be specified in command line.\n'
print usage
sys.exit()
non_null_pklpath = args[0]
null_pklpath = args[1]
# parse ops
score_threshold = float(options.score_threshold)
# get dics from pkl
non_null_dic = get_dic_from_pklpath(non_null_pklpath)
null_dic = get_dic_from_pklpath(null_pklpath)
non_null_gerp_scores = get_gerp_scores(non_null_dic,
gerpkey='avg_rs_score')
null_gerp_scores = get_gerp_scores(null_dic, gerpkey='avg_rs_score')
plot_functions.plot_density([non_null_gerp_scores, null_gerp_scores],
mytitle='Density plot of conservation scores',
labels_lists=['MEME motifs', 'Controls'],
xlabel='GERP conservation score',
ylabel='Density',
xmin=-4,
xmax=4,
ymax=options.ymax,
smoothness=0.15,
drawvline=score_threshold)
# find how many conserved regions are in each.
n_conserved_in_meme = \
gerp_utilities.conserved_regions(non_null_gerp_scores, fraction=False, threshold=score_threshold)
n_conserved_in_null = \
gerp_utilities.conserved_regions(null_gerp_scores, fraction=False, threshold=score_threshold)
n_total_in_meme = len(non_null_gerp_scores)
n_total_in_null = len(null_gerp_scores)
n_not_conserved_in_meme = n_total_in_meme - n_conserved_in_meme
n_not_conserved_in_null = n_total_in_null - n_conserved_in_null
print 'Threshold: %s' % score_threshold
print 'Number of conserved elements: %s' % n_conserved_in_meme
print 'Number of conserved elements found in control: %s' % n_conserved_in_null
# Perform fisher's exact test
oddsratio, pvalue = fisher_exact(
[[n_conserved_in_meme, n_conserved_in_null],
[n_not_conserved_in_meme, n_not_conserved_in_null]])
print 'Fishers Exact Test, Oddsratio: %s. Pvalue: %s' % (oddsratio, pvalue)
# plot distributions
mylabels = ['Meme motifs', 'Control region']
mytitle = 'Fraction of elements conserved compared to control region'
# Plot bargraphs
frac_conserved_meme = float(n_conserved_in_meme) / n_total_in_meme
frac_conserved_null = float(n_conserved_in_null) / n_total_in_null
myvals = [frac_conserved_meme, frac_conserved_null]
plot_functions.plot_barplot(myvals, mytitle, mylabels,
ylabel='Fraction of elements conserved',
mytext1="%i/%i" \
%(n_conserved_in_meme,
n_total_in_meme),
mytext2='%i/%i' %(n_conserved_in_null,
n_total_in_null),
mytext3="*Fisher's Exact Test\nP-value=%.2e" %pvalue,
ymin=0,
ymax=1,
width=0.5)
plt.show()
