1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325
|
import numpy as np
import matplotlib
matplotlib.use('Agg')
import pylab as pl
import argparse
import os
from sklearn.metrics import auc
from matplotlib.font_manager import FontProperties
# find bounds for plotting
# threshold tells, how many reads are to be ignored on both sides (relative to length of the list)
def find_bounds(scorelist, threshold=.0005):
all_length = len(scorelist)
i = 0
j = all_length
scorelist.sort()
while i < threshold * all_length:
i += 1
while all_length - j < threshold * all_length:
j -= 1
return scorelist[i][0], scorelist[j - 1][0]
def find_bounds_unassigned(scorelist, threshold=.0005):
all_length = len(scorelist)
i = 0
j = all_length
scorelist.sort()
while i < threshold * all_length:
i += 1
while all_length - j < threshold * all_length:
j -= 1
return float(scorelist[i]), float(scorelist[j - 1])
def hist_of_scores(species_score, path, name1, name2, k, amount, SE_String):
species1 = []
species2 = []
bins2 = [0] * 101
bins3 = [0] * 101
bounds = find_bounds(species_score)
lowerbound = bounds[0]
upperbound = bounds[1]
stepsize = (bounds[1] - bounds[0]) / 101.
for key in species_score:
logkey = (key[0])
if key[1] == '1':
species1.append(logkey)
elif key[1] == '2':
species2.append(logkey)
for i in range(len(species1)):
if lowerbound < species1[i] < upperbound:
bins2[int((species1[i] - lowerbound) / stepsize)] += 1
for i in range(len(species2)):
if lowerbound < species2[i] < upperbound:
bins3[int((species2[i] - lowerbound) / stepsize)] += 1
ticks = np.arange(bounds[0], bounds[1] + bounds[1] / 1000., (bounds[1] - bounds[0]) / 100.)
pl.plot(ticks, bins2, label=name1, color='blue')
pl.plot(ticks, bins3, label=name2, color='red')
pl.legend(loc='upper right', numpoints=1)
pl.savefig(os.path.join(path, 'score_histogram_k' + str(k) + '_' + str(amount) + SE_String + '.png'))
return species1, species2, bounds
def fasta_generator(infile):
block = []
state = 'name'
for line in infile:
if state == 'name':
block.append(line.rstrip())
state = 'read'
elif state == 'read':
block.append(line.rstrip())
state = 'name'
yield block
block = []
def fastq_generator(infile):
block = []
qual = ''
read = ''
state = 'name' # name read qual
counter = 0
for line in infile:
if state == 'name':
block.append(line.rstrip())
state = 'read'
elif state == 'read':
if line[0] == '+':
block.append(read)
block.append('+')
read = ''
state = 'qual'
else:
counter += 1
read += line.rstrip()
elif state == 'qual':
if counter > 1:
qual += line.rstrip()
counter -= 1
else:
qual += line.rstrip()
counter = 0
block.append(qual)
qual = ''
state = 'name'
yield block
block = []
def hist_distribution_plot(infile, path, species_hist_scores, name1, name2, k, amount, SE_String, filetype):
if filetype == 'fastq':
gen = fastq_generator(open(infile))
else:
gen = fasta_generator(open(infile))
block = next(gen, False)
species_by_cutoff = []
while block:
species_by_cutoff.append((float(block[0].split('_')[-2]) - float(block[0].split('_')[-1])))
block = next(gen, False)
bounds = find_bounds_unassigned(species_by_cutoff)
lowerbound = min(species_hist_scores[2][0], bounds[0])
upperbound = max(species_hist_scores[2][1], bounds[1])
species1_scores = species_hist_scores[0]
species2_scores = species_hist_scores[1]
bins = [0] * 101
bins2 = [0] * 101
bins3 = [0] * 101
stepsize = (upperbound - lowerbound) / 100.
for logkey in species_by_cutoff:
if logkey > lowerbound and logkey < upperbound:
bins[int((logkey - lowerbound) / stepsize)] += 1
for i in range(min(len(species1_scores), len(species2_scores))):
if species1_scores[i] > lowerbound and species1_scores[i] < upperbound:
bins2[int((species1_scores[i] - lowerbound) / stepsize)] += 1
if species2_scores[i] > lowerbound and species2_scores[i] < upperbound:
bins3[int((species2_scores[i] - lowerbound) / stepsize)] += 1
sum_bins = float(sum(bins))
for each in range(len(bins)):
bins[each] = (bins[each] / sum_bins) # +1 to get positive logarithms
a = []
b = bins
sum_bins2 = float(sum(bins2))
sum_bins3 = float(sum(bins3))
for each in range(len(bins)):
bins2[each] /= sum_bins2
bins3[each] /= sum_bins3
for each in range(len(bins2)):
a.append([bins2[each], bins3[each]])
x, y = np.linalg.lstsq(a, b)[0]
xplusy = x + y
x = x / xplusy
y = y / xplusy
sum_of_hist_scores = []
for each in range(len(bins2)):
bins[each] += 1
bins2[each] = (max(0, bins2[each] * x)) + 1
bins3[each] = (max(0, bins3[each] * y)) + 1 # +1 to get positive logarithms
sum_of_hist_scores.append(bins2[each] * bins3[each])
ticks = np.arange(lowerbound, upperbound + upperbound / 1000., (upperbound - lowerbound) / 100.)
fig = pl.figure()
fig.set_size_inches(12, 9.5)
ax = fig.add_subplot(1, 1, 1)
ax.plot(ticks, bins, label='real data', color='purple', linewidth=3, alpha=1)
ax.plot(ticks, bins2, label='simulated data of ' + name1, color='blue', linewidth=3, alpha=.25)
ax.fill_between(ticks, bins2, 1, color='blue', alpha=0.25, offset_position='screen')
ax.plot(ticks, bins3, label='simulated data of ' + name2, color='red', linewidth=3, alpha=.25)
ax.fill_between(ticks, bins3, 1, color='red', alpha=0.25, offset_position='screen')
ax.plot(ticks, sum_of_hist_scores, label='sum of simulated data', color='green', linewidth=3, alpha=1)
ax.set_yscale('log')
ax.axes.get_yaxis().set_visible(False)
fontp = FontProperties()
fontp.set_size(27)
pl.title(name1 + ' (' + str(round(x * 100, 2)) + '%) and ' + name2 + ' (' + str(round(y * 100, 2)) + '%) data',
fontsize=40.)
pl.ylabel('Frequency', size=40.)
ax.set_yticks([0, 0.5, 1])
pl.xlabel('Read score', size=40.)
pl.ylim(0, 1.02 * (
max((max(bins), max(bins2), max(bins3))) + (max((max(bins), max(bins2),
max(bins3)))) / 500.)) # normalize height
leg = pl.legend(loc='upper right', numpoints=1, prop=fontp)
for item in ([ax.title, ax.xaxis.label, ax.yaxis.label]):
item.set_fontsize(27)
for item in (ax.get_xticklabels() + ax.get_yticklabels()):
item.set_fontsize(20)
ax.text(-.01, .005, '0', horizontalalignment='right', verticalalignment='center', rotation='horizontal',
transform=ax.transAxes, size=20)
ax.text(-.01, .475, '.5', horizontalalignment='right', verticalalignment='center', rotation='horizontal',
transform=ax.transAxes, size=20)
ax.text(-.01, .95, '1', horizontalalignment='right', verticalalignment='center', rotation='horizontal',
transform=ax.transAxes, size=20)
ax.text(-.04, .5, 'Frequency [rel. units]', horizontalalignment='right', verticalalignment='center',
rotation='vertical', transform=ax.transAxes, size=27)
leg.get_frame().set_alpha(0.5)
pl.savefig(os.path.join(path, 'fitted_histograms_k' + str(k) + '_' + str(amount) + SE_String + '.png'))
def parse_fasta(infile, name1, name2):
infile = open(infile)
species_by_cutoff = []
name = infile.readline()
infile.readline()
while name[0] == '>':
name = name[::-1].split('_', 2)[::-1]
name = [x[::-1].rstrip() for x in name]
if name[0][1:len(name1) + 1] == name1:
species_by_cutoff.append([float(name[-2]) - float(name[-1]), '1'])
elif name[0][1:len(name2) + 1] == name2:
species_by_cutoff.append([float(name[-2]) - float(name[-1]), '2'])
else:
raise Exception(
'There seems to be a problem with the species names. Try to avoid special characters or use the default values')
name = infile.readline()
infile.readline()
if name == '':
name = 'ending now'
return species_by_cutoff
def parse_fastq(infile, name1, name2):
infile = open(infile)
species_by_cutoff = []
name = '@'
while name[0][0] == '@':
name = infile.readline()
infile.readline()
infile.readline()
infile.readline()
if len(name) <= 1:
name = 'ending now'
name = name.split('_')
if name[0][1] == name1[0]:
species_by_cutoff.append([float(name[-2]) - float(name[-1]), '1'])
elif name[0][1] == name2[0]:
species_by_cutoff.append([float(name[-2]) - float(name[-1]), '2'])
return species_by_cutoff
def roc_plot(species_by_cutoff, k, path, amount, SE_String):
tpr = [0]
fpr = [0]
species_by_cutoff = sorted(species_by_cutoff)
tpl = []
fontp = FontProperties()
fontp.set_size(27)
for each in species_by_cutoff:
tpl.append(each[1])
species1count = 0
species2count = 0
for i in range(len(tpl)):
if tpl[i] == '2':
species1count += 1
elif tpl[i] == '1':
species2count += 1
tpr.append(species1count / float(
len(species_by_cutoff) / 2.)) # this will not work if both species appear differently often!
fpr.append(species2count / float(len(species_by_cutoff) / 2.))
roc_auc = auc(fpr, tpr)
# Plot ROC curve
pl.clf()
fig = pl.figure()
fig.set_size_inches(12, 9.5)
ax = fig.add_subplot(1, 1, 1)
ax.plot(fpr, tpr, label='ROC curve (area = %0.2f)' % roc_auc, color='green', zorder=10, lw=4, alpha=.6)
ax.plot([0.0, 1], [0, 1], 'k--')
ax.fill_between(fpr, tpr, color=(44 / 255., 160 / 255., 44 / 255.), alpha=.6)
pl.xlabel('False Positive Rate', fontsize=27)
pl.ylabel('True Positive Rate', fontsize=27)
pl.title('ROC-Curve of ' + str(k) + '-mer based read distinction', fontsize=27)
pl.legend(loc='lower right', prop=fontp)
for item in (ax.get_xticklabels() + ax.get_yticklabels()):
item.set_fontsize(20)
pl.savefig(os.path.join(path, 'ROCplot_k' + str(k) + '_' + str(amount) + SE_String + '.png'))
def main(k, path, name1, name2, infile_simulated, infile_real, amount, SE, filetype='fastq'):
all_return_vals = parse_fasta(infile_simulated, name1, name2)
species_by_cutoff = all_return_vals
if SE:
se_string = '_SE'
else:
se_string = '_PE'
print('Plotting histogram of scores')
species_scores = hist_of_scores(species_by_cutoff, path, name1, name2, k, amount, se_string)
pl.close('all')
print('Plotting ROC-plot')
roc_plot(species_by_cutoff, k, path, amount, se_string)
pl.close('all')
print('Plotting distribution heights')
hist_distribution_plot(infile_real, path, species_scores, name1, name2, k, amount, se_string, filetype)
pl.close('all')
if __name__ == '__main__':
parser = argparse.ArgumentParser()
parser.add_argument('-k', '--kmer', required=True)
parser.add_argument('-p', '--path', required=True)
parser.add_argument('-a', '--amount', required=False, default=50000)
parser.add_argument('-n', '--names', required=True, action='append',
help='give 2 species names (-n SPECIES1 -n SPECIES2)')
parser.add_argument('-t', '--filetype', default='fastq',
help='specify whether your input is fastq of fasta-type. Default = fastq. Type -t fasta to change to fasta',
required=False)
parser.add_argument('-s', 'simulated', required=True)
parser.add_argument('-r', 'real', required=True)
args = parser.parse_args()
path = args.path
amount = int(args.amount)
filetype = args.filetype
name1 = args.names[0]
name2 = args.names[1]
k = args.kmer
infile_simulated = str(args.simulated)
infile_real = str(args.real)
main(k, path, name1, name2, infile_simulated, infile_real, amount, filetype)
|