def _compute_skews(self):
### initialisation = Calculating GC skew and AT skew for first window
self._init_sliding_window()
GC_content_slide, GC_skew_slide = self._init_list_results()
AT_content_slide, AT_skew_slide = self._init_list_results()
self._init_cumul_nuc()
c = Counter(self._slide_window)
dict_counts = {'G': c['G'], 'C': c['C'], 'A': c['A'], 'T': c['T']}
i = 0
# GC
sumGC = float(dict_counts['G'] + dict_counts['C'])
GC_content_slide[0][i] = sumGC
if sumGC > 0:
GC_skew_slide[0][i] = (dict_counts['G'] - dict_counts['C']) / sumGC
# AT
sumAT = float(dict_counts['A'] + dict_counts['T'])
AT_content_slide[0][i] = sumAT
if sumAT > 0:
AT_skew_slide[0][i] = (dict_counts['A'] - dict_counts['T']) / sumAT
### Compute for all genome
while (self._seq_right):
out_nuc = self._slide_window.popleft()
in_nuc = self._seq_right.popleft()
self._slide_window.append(in_nuc)
i += 1
if i % 500000 == 0:
logger.info("%d / %d" % (i, self.__len__()))
# if in and out are the same : do nothing, append same result
if out_nuc != in_nuc:
# remove out from counters
if out_nuc in self._dict_nuc:
dict_counts[out_nuc] -= 1
if in_nuc in self._dict_nuc:
dict_counts[in_nuc] += 1
sumGC = float(dict_counts['G'] + dict_counts['C'])
sumAT = float(dict_counts['A'] + dict_counts['T'])
# fill results
# GC
GC_content_slide[0][i] = sumGC
if sumGC > 0:
GC_skew_slide[0][i] = (dict_counts['G'] -
dict_counts['C']) / sumGC
# AT
AT_content_slide[0][i] = sumAT
if sumAT > 0:
AT_skew_slide[0][i] = (dict_counts['A'] -
dict_counts['T']) / sumAT
# cumul
if in_nuc in self._dict_nuc:
self._cumul[self._dict_nuc[in_nuc]][i + self._window - 1] += 1
self._GC_content_slide = GC_content_slide / float(self._window)
self._AT_content_slide = AT_content_slide / float(self._window)
self._cumul = np.delete(self._cumul,
range(self.__len__(), self._cumul.shape[1]), 1)
self._cumul = np.cumsum(self._cumul, axis=1)
### save result for Z curve
self._Xn = list((self._cumul[self._dict_nuc['A']] + self._cumul[self._dict_nuc['G']]) -\
(self._cumul[self._dict_nuc['C']] + self._cumul[self._dict_nuc['T']]))
self._Yn = list((self._cumul[self._dict_nuc['A']] + self._cumul[self._dict_nuc['C']]) -\
(self._cumul[self._dict_nuc['G']] + self._cumul[self._dict_nuc['T']]))
self._Zn = list((self._cumul[self._dict_nuc['A']] + self._cumul[self._dict_nuc['T']]) -\
(self._cumul[self._dict_nuc['C']] + self._cumul[self._dict_nuc['G']]))
self._AT_skew_slide = AT_skew_slide
self._GC_skew_slide = GC_skew_slide
### check proportion of ignored nucleotides
GC_content_total = (self._cumul[self._dict_nuc['G']][-1] +
self._cumul[self._dict_nuc['C']][-1]) / float(
self.__len__())
AT_content_total = (self._cumul[self._dict_nuc['A']][-1] +
self._cumul[self._dict_nuc['T']][-1]) / float(
self.__len__())
self._ignored_nuc = 1.0 - GC_content_total - AT_content_total
def _compute_skews(self):
### initialisation = Calculating GC skew and AT skew for first window
self._init_sliding_window()
GC_content_slide, GC_skew_slide = self._init_list_results()
AT_content_slide, AT_skew_slide = self._init_list_results()
self._init_cumul_nuc()
c = Counter(self._slide_window)
dict_counts = {'G' : c['G'], 'C' : c['C'], 'A' : c['A'], 'T' : c['T']}
i = 0
# GC
sumGC = float(dict_counts['G'] + dict_counts['C'])
GC_content_slide[0][i] = sumGC
if sumGC > 0:
GC_skew_slide[0][i] = (dict_counts['G'] - dict_counts['C']) / sumGC
# AT
sumAT = float(dict_counts['A'] + dict_counts['T'])
AT_content_slide[0][i] = sumAT
if sumAT > 0:
AT_skew_slide[0][i] = (dict_counts['A'] - dict_counts['T']) / sumAT
### Compute for all genome
while(self._seq_right):
out_nuc = self._slide_window.popleft()
in_nuc = self._seq_right.popleft()
self._slide_window.append(in_nuc)
i += 1
if i % 500000 == 0:
logger.info("%d / %d" % (i, self.__len__()))
# if in and out are the same : do nothing, append same result
if out_nuc != in_nuc:
# remove out from counters
if out_nuc in self._dict_nuc:
dict_counts[out_nuc] -= 1
if in_nuc in self._dict_nuc:
dict_counts[in_nuc] += 1
sumGC = float(dict_counts['G'] + dict_counts['C'])
sumAT = float(dict_counts['A'] + dict_counts['T'])
# fill results
# GC
GC_content_slide[0][i] = sumGC
if sumGC > 0:
GC_skew_slide[0][i] = (dict_counts['G'] - dict_counts['C'])/sumGC
# AT
AT_content_slide[0][i] = sumAT
if sumAT > 0:
AT_skew_slide[0][i] = (dict_counts['A'] - dict_counts['T'])/sumAT
# cumul
if in_nuc in self._dict_nuc:
self._cumul[self._dict_nuc[in_nuc]][i+self._window-1] +=1
self._GC_content_slide = GC_content_slide/float(self._window)
self._AT_content_slide = AT_content_slide/float(self._window)
self._cumul = np.delete(self._cumul, range(self.__len__(),self._cumul.shape[1]),1)
self._cumul = np.cumsum(self._cumul,axis=1)
### save result for Z curve
self._Xn = list((self._cumul[self._dict_nuc['A']] + self._cumul[self._dict_nuc['G']]) -\
(self._cumul[self._dict_nuc['C']] + self._cumul[self._dict_nuc['T']]))
self._Yn = list((self._cumul[self._dict_nuc['A']] + self._cumul[self._dict_nuc['C']]) -\
(self._cumul[self._dict_nuc['G']] + self._cumul[self._dict_nuc['T']]))
self._Zn = list((self._cumul[self._dict_nuc['A']] + self._cumul[self._dict_nuc['T']]) -\
(self._cumul[self._dict_nuc['C']] + self._cumul[self._dict_nuc['G']]))
self._AT_skew_slide = AT_skew_slide
self._GC_skew_slide = GC_skew_slide
### check proportion of ignored nucleotides
GC_content_total = (self._cumul[self._dict_nuc['G']][-1] +
self._cumul[self._dict_nuc['C']][-1]) / float(self.__len__())
AT_content_total = (self._cumul[self._dict_nuc['A']][-1] +
self._cumul[self._dict_nuc['T']][-1]) / float(self.__len__())
self._ignored_nuc = 1.0 - GC_content_total - AT_content_total