class DeltaExtractor(object):
def __init__(self, path):
self.path = path
self.masterSpectrum = MasterSpectrum()
def createDeltaMasterSpectrum(self,
min_rel_intensity,
delta_func=calculate_Delta_by_ppm(20)):
up = 0
with mgf.read(self.path) as spectra:
for spectrum in spectra:
print(up)
up += 1
int_dic = spectrum['intensity array']
mz_dic = spectrum['m/z array']
rel_int = calculateRelativeIntensity(int_dic)
smallerArea = [(i, j) for i, j in zip(mz_dic, rel_int)
if j >= min_rel_intensity]
mz_dic = [i for i, j in smallerArea]
rel_int = [j for i, j in smallerArea]
for i in range(len(mz_dic) - 1, -1, -1):
for j in range(i - 1, -1, -1):
diff = mz_dic[i] - mz_dic[j]
p = Peak(diff, rel_int[j],
delta_func) # intensities are from lower peak
self.masterSpectrum.add(p, 0)
def exportCsv(self, output_path):
self.masterSpectrum.export_to_csv(output_path)
def __init__(self, path, output_path, min_rel_similarity, ppm=10):
self.ppm = ppm
self.ms = MasterSpectrum()
self.path = path
self.out = []
self.output_path = output_path
self.references = sortedlist(key=lambda i: i.id_1)
self.min_rel_similarity = min_rel_similarity
def request_ms(self):
ms = MasterSpectrum()
delta_func = calculate_Delta_by_ppm(20)
for m in self.mz_ary:
p = Peak(float(m), 1, delta_func)
ms.add(p, 0)
return ms
def load_recalibrate(self):
fc = calculate_Delta_by_ppm(self.ppm)
tmt_mass = calculate_tag_tmt10()
with mgf.read(self.path) as spectra:
for spectrum in spectra:
ms = MasterSpectrum()
params = spectrum['params']
for mass, intensity in zip(spectrum['m/z array'],
spectrum['intensity array']):
ms.add(Peak(mass, intensity, fc))
peak = Peak(tmt_mass, 0.5, fc)
if peak.key() not in ms.spectrum[0]:
recalibrate = False
else:
idx, bin_to_ack, a, b = ms.binary(
peak, 0,
len(ms.spectrum[0][peak.key()]) - 1, 0)
if idx == -1:
recalibrate = False
else:
recalibrate = True
recalibration_mass = ms.spectrum[0][peak.key()][idx].mz
diff = tmt_mass - recalibration_mass
print(params['title'])
print("original={0}\tdiff={1}".format(
recalibration_mass, diff))
mass_list = []
int_list = []
if recalibrate:
ppm_shift = calculate_ppm_shift(diff, tmt_mass)
for key in ms.spectrum[0].keys():
for mp in ms.spectrum[0][key]:
if recalibrate:
if self.type == 'ppm':
diff = calculate_da_shift(mp.mz, ppm_shift)
mass_list.append(mp.mz + diff)
elif self.type == 'absolute':
diff = diff
mass_list.append(mp.mz + diff)
else:
print(self.type)
raise ValueError("what did you dooooo")
else:
mass_list.append(mp.mz)
int_list.append(mp.intensity)
print("len is:\t{0}".format(len(mass_list)))
mgf.write(spectra=[{
'm/z array': mass_list,
'intensity array': int_list,
'params': params
}],
output=self.file_out)
def test_weird_summing_mgf(self):
path = "tests/data/weird_sum.mgf"
output = "/home/tobiass/weird_sum.csv"
output_mgf = "tests/data/temp/weird_mgf.mgf"
ma = MgfAnnotater(path, output)
ma.load_msconvert_mgf()
ma.export_annotated_spectra_to_mgf(output_mgf)
ms = MasterSpectrum()
ms.load_from_mgf("tests/data/temp/weird_mgf.mgf", ignoreCharges=True)
assert_equal(round(ms.spectrum[0][179][0].intensity, 4), round(0.0025, 4))
class DeltaPatcher(object):
def __init__(self, delta_func=calculate_Delta_by_ppm(20)):
self.exclusionSpectrum = MasterSpectrum()
self.delta_function = delta_func
self.precursorDeltas = []
def readExclusionList(self, path):
'''
exclusionList: 3 columns
m/z , comments
'''
with open(path, 'r') as csvfile:
readr = csv.reader(csvfile)
header = True
for row in readr:
if header:
header = False
else:
if row[2] == 'absolute':
p = Peak(float(row[0]), 1.0, self.delta_function)
mp = MasterPeak(p)
# 0 for no differentiation of charge states
self.exclusionSpectrum.add(mp, 0)
elif row[2] == 'precursor':
self.precursorDeltas.append(float(row[0]))
else:
raise ValueError('A very specific bad thing happened')
def patchDelta(self, input_path, output_path):
'''
'''
with open(output_path, "wt") as csvfile:
writr = csv.writer(csvfile, lineterminator=os.linesep)
with open(input_path, 'r') as r_csvfile:
readr = csv.reader(r_csvfile)
header = True
for row in readr:
if header:
header = False
writr.writerow(row)
else:
peak = Peak(float(row[0]), 0, self.delta_function)
if peak.key() in self.exclusionSpectrum.spectrum[0]:
idx, bin_to_ack, should_merge_left_peak, should_merge_right_peak = self.exclusionSpectrum.binary(
peak, 0,
len(self.exclusionSpectrum.spectrum[0][
peak.key()]) - 1, 0)
if idx != -1: # found
print("found it")
else: # not found
writr.writerow(row)
else:
writr.writerow(row)
def just_create_heavy_ms(self):
ms = MasterSpectrum()
for mass, i in zip(self.peak_list_1, self.int_list_1):
ms.add(
Peak(mass,
i,
self.fc,
meta={
'delta': 511,
'mass': -1,
'decharged': False
}))
return ms
def generateMS_by_score_file(score_info_object):
"""
every object has an mz array
every object has an diff array
"""
ms = MasterSpectrum()
delta_func = calculate_Delta_by_ppm(20)
dPeaks_matched = 0
for m, diff in zip(score_info_object['mz'], score_info_object['diff']):
p = Peak(float(m), 1, delta_func, meta=diff)
ms.add(p, 0)
dPeaks_matched += 1
return ms, dPeaks_matched
class PrecursorDeltaPeaksExtractor(object):
def __init__(self, path):
self.path = path
self.masterSpectrum = MasterSpectrum()
def createDeltaPrecursorMasterSpectrum(
self, delta_func=calculate_Delta_by_ppm(20)):
with mgf.read(self.path) as spectra:
for spectrum in spectra:
int_dic = spectrum['intensity array']
mz_dic = spectrum['m/z array']
chrg_spec = spectrum['params']['charge'][0]
precursor = calculatePrecursor(
mz=spectrum['params']['pepmass'][0], charge=chrg_spec)
rel_int = calculateRelativeIntensity(int_dic)
for m, i in zip(mz_dic, rel_int):
p = Peak(precursor - float(m), float(i), delta_func)
self.masterSpectrum.add(p, 0)
def exportCsv(self, output_path):
self.masterSpectrum.export_to_csv(output_path)
def test_patch_mgf(self):
ms = MasterSpectrum()
ms.load_from_mgf(
'tests/data/cetsa_101287_A01_P013190_S00_N01_R1_TMT10_3spectra_shorted.mgf',
ignoreCharges=True)
assert_equal(177 in ms.spectrum[0].keys(), True)
mgfP = MgfPatcher(delta_func=calculate_Delta_by_ppm(20))
mgfP.readExclusionList('tests/data/exclusionList.txt')
mgfP.patchMgf(
'tests/data/cetsa_101287_A01_P013190_S00_N01_R1_TMT10_3spectra_shorted.mgf',
'tests/data/temp/cetsa_101287_A01_P013190_S00_N01_R1_TMT10_3spectra_shorted_patched.mgf'
)
ms = MasterSpectrum()
ms.load_from_mgf(
'tests/data/temp/cetsa_101287_A01_P013190_S00_N01_R1_TMT10_3spectra_shorted_patched.mgf',
ignoreCharges=True)
assert_equal(177 in ms.spectrum[0].keys(), False)
assert_equal(433 in ms.spectrum[0].keys(), False)
def load_distiller_mgf2(self):
"""
creates references based on precursor mass
a missing scanid means an ms1 event
by default referencing works just within one ms2 block
"""
fc = calculate_Delta_by_ppm(self.ppm)
error = 0
self.ms = MasterSpectrum()
with mgf.read(self.path) as spectra:
for spectrum in spectra:
mass = spectrum['params']['pepmass'][0]
precursor_chrg = int(spectrum['params']['charge'][0])
mass = calculatePrecursor(mass, precursor_chrg)
scanid = int(parse_scan_id(spectrum['params']['title']))
found = False
if len(self.ms.spectrum) == 0:
peak = Peak(mass, scanid, fc)
self.ms.add(Peak(mass, scanid, fc, meta={'ms': spectrum['m/z array'], 'int': spectrum['intensity array'], 'params': spectrum['params']}),
charge=precursor_chrg)
found = True
else:
if (precursor_chrg in self.ms.spectrum.keys()): # react to charge !!!!!!
if len(self.ms.spectrum[precursor_chrg]) == 0:
peak = Peak(mass, scanid, fc)
self.ms.add(Peak(mass, scanid, fc, meta={'ms': spectrum['m/z array'], 'int': spectrum['intensity array'], 'params': spectrum['params']}),
charge=precursor_chrg)
found = True
else:
for extra_mass in gen_allowed_mass_diff_with_sign(n=4, z=1):
if found is False:
peak = Peak(mass + extra_mass, 0.5, fc)
if peak.key() in self.ms.spectrum[precursor_chrg]:
print(precursor_chrg)
idx, bin_to_ack, a, b = self.ms.binary(peak, 0, len(self.ms.spectrum[precursor_chrg][peak.key()]) - 1, precursor_chrg)
if idx != -1:
self.references.add(Reference(ppm=self.ppm,
id_2=scanid,
id_1=self.ms.spectrum[precursor_chrg][peak.key()][idx].intensity, # also scanid
peak_list_2=spectrum['m/z array'],
peak_list_1=self.ms.spectrum[precursor_chrg][peak.key()][idx].meta['ms'],
mass_2=mass,
mass_1=self.ms.spectrum[precursor_chrg][peak.key()][idx].mz,
charge=spectrum['params']['charge'][0],
extra_mass=extra_mass,
int_list_2=spectrum['intensity array'],
int_list_1=self.ms.spectrum[precursor_chrg][peak.key()][idx].meta['int'],
params2=spectrum['params'],
params1=self.ms.spectrum[precursor_chrg][peak.key()][idx].meta['params']))
found = True
del(self.ms.spectrum[precursor_chrg][peak.key()][idx])
if len(self.ms.spectrum[precursor_chrg][peak.key()]) == 0:
del(self.ms.spectrum[precursor_chrg][peak.key()])
if len(self.ms.spectrum[precursor_chrg]) == 0:
del(self.ms.spectrum[precursor_chrg])
if found is False:
limit_scan_id = scanid - 20 # could start at -19
ms_bac = MasterSpectrum()
for chrg in self.ms.spectrum:
for key in self.ms.spectrum[chrg].keys():
for mp in self.ms.spectrum[chrg][key]:
if mp.intensity >= limit_scan_id:
ms_bac.add(mp, charge=chrg)
self.ms = ms_bac
self.ms.add(Peak(mass, scanid, fc, meta={'ms': spectrum['m/z array'], 'int': spectrum['intensity array'], 'params': spectrum['params']}),
charge=precursor_chrg)
if error > 0:
print(" delete valid information {0}".format(error))
class MgfPatcher(object):
def __init__(self, delta_func=calculate_Delta_by_ppm(20)):
self.exclusionSpectrum = MasterSpectrum()
self.delta_function = delta_func
self.precursorDeltas = []
def readExclusionList(self, path):
'''
exclusionList: 2 columns
m/z , comments
'''
with open(path, 'r') as csvfile:
readr = csv.reader(csvfile)
header = True
for row in readr:
if header:
header = False
else:
if row[2] == 'absolute':
p = Peak(float(row[0]), 1.0, self.delta_function)
mp = MasterPeak(p)
# 0 for no differentiation of charge states
self.exclusionSpectrum.add(mp, 0)
elif row[2] == 'precursor':
self.precursorDeltas.append(float(row[0]))
else:
raise ValueError('A very specific bad thing happened')
def patchMgf(self, input_path, output_path):
'''
'''
with mgf.read(input_path) as spectra:
spectra_out = []
for spectrum in spectra:
int_dic = spectrum['intensity array']
mz_dic = spectrum['m/z array']
param_dic = spectrum['params']
chrg_spec = spectrum['params']['charge'][0]
precursor = calculatePrecursor(
mz=spectrum['params']['pepmass'][0], charge=chrg_spec)
pos = 0
del_array = []
for m in mz_dic:
peak = Peak(m, 0, self.delta_function)
if peak.key() in self.exclusionSpectrum.spectrum[0]:
idx, bin_to_ack, should_merge_left_peak, should_merge_right_peak = self.exclusionSpectrum.binary(
peak, 0,
len(self.exclusionSpectrum.spectrum[0][peak.key()])
- 1, 0)
if idx != -1: # found
del_array.append(pos)
else:
mp = MasterPeak(peak)
for precursorDelta in self.precursorDeltas:
if mp.isInsideMz(precursor - precursorDelta):
del_array.append(pos)
else:
pass
pos += 1
int_dic = np.delete(int_dic, del_array, 0)
mz_dic = np.delete(mz_dic, del_array, 0)
spectra_out.append({
'm/z array': mz_dic,
'intensity array': int_dic,
'params': param_dic
})
mgf.write(spectra=spectra_out, output=output_path)
def __init__(self, delta_func=calculate_Delta_by_ppm(20)):
self.exclusionSpectrum = MasterSpectrum()
self.delta_function = delta_func
self.precursorDeltas = []
class MgfAnnotater(object):
def __init__(self, path, output_path, min_rel_similarity, ppm=10):
self.ppm = ppm
self.ms = MasterSpectrum()
self.path = path
self.out = []
self.output_path = output_path
self.references = sortedlist(key=lambda i: i.id_1)
self.min_rel_similarity = min_rel_similarity
def load_msconvert_mgf(self):
"""
creates references based on precursor mass
a missing scanid means an ms1 event
by default referencing works just within one ms2 block
"""
fc = calculate_Delta_by_ppm(self.ppm)
scan_id_ary = []
problems = []
error = 0
with mgf.read(self.path) as spectra:
for spectrum in spectra:
mass = spectrum['params']['pepmass'][0]
precursor_chrg = int(spectrum['params']['charge'][0])
mass = calculatePrecursor(mass, precursor_chrg)
scanid = int(parse_scan_id(spectrum['params']['title']))
if len(scan_id_ary) == 0:
scan_id_ary.append(scanid)
else:
if scanid != scan_id_ary[-1] + 1:
if len(scan_id_ary) % 2 == 1:
problems.append(scan_id_ary[0])
error += 1
scan_id_ary = []
scan_id_ary.append(scanid)
else:
scan_id_ary = []
scan_id_ary.append(scanid)
self.ms = MasterSpectrum(
) # new MS if scan_id group (seperated by ms1) is completed
else:
scan_id_ary.append(scanid)
found = False
if len(self.ms.spectrum) == 0:
peak = Peak(mass, scanid, fc)
self.ms.add(Peak(mass,
scanid,
fc,
meta={
'ms': spectrum['m/z array'],
'int': spectrum['intensity array'],
'params': spectrum['params']
}),
charge=precursor_chrg)
found = True
else:
if (precursor_chrg in self.ms.spectrum.keys()
): # react to charge !!!!!!
if len(self.ms.spectrum[precursor_chrg]) == 0:
peak = Peak(mass, scanid, fc)
self.ms.add(Peak(mass,
scanid,
fc,
meta={
'ms': spectrum['m/z array'],
'int':
spectrum['intensity array'],
'params': spectrum['params']
}),
charge=precursor_chrg)
found = True
else:
for extra_mass in gen_allowed_mass_diff_with_sign(
n=4, z=1):
if found is False:
peak = Peak(mass + extra_mass, 0.5, fc)
if peak.key(
) in self.ms.spectrum[precursor_chrg]:
print(precursor_chrg)
idx, bin_to_ack, a, b = self.ms.binary(
peak, 0,
len(self.ms.spectrum[
precursor_chrg][peak.key()]) -
1, precursor_chrg)
if idx != -1:
self.references.add(
Reference(
ppm=self.ppm,
id_2=scanid,
id_1=self.ms.
spectrum[precursor_chrg][
peak.key()][idx].
intensity, # also scanid
peak_list_2=spectrum[
'm/z array'],
peak_list_1=self.ms.
spectrum[precursor_chrg][
peak.key()]
[idx].meta['ms'],
mass_2=mass,
mass_1=self.ms.
spectrum[precursor_chrg][
peak.key()][idx].mz,
charge=spectrum['params']
['charge'][0],
extra_mass=extra_mass,
int_list_2=spectrum[
'intensity array'],
int_list_1=self.ms.
spectrum[precursor_chrg][
peak.key()]
[idx].meta['int'],
params2=spectrum['params'],
params1=self.ms.
spectrum[precursor_chrg][
peak.key()]
[idx].meta['params']))
found = True
del (self.ms.
spectrum[precursor_chrg][
peak.key()][idx])
if len(self.ms.
spectrum[precursor_chrg][
peak.key()]) == 0:
del (self.ms.
spectrum[precursor_chrg][
peak.key()])
if len(self.ms.spectrum[
precursor_chrg]) == 0:
del (self.ms.spectrum[
precursor_chrg])
if found is False:
self.ms.add(Peak(mass,
scanid,
fc,
meta={
'ms': spectrum['m/z array'],
'int': spectrum['intensity array'],
'params': spectrum['params']
}),
charge=precursor_chrg)
if error > 0:
print(" delete valid information {0}".format(error))
# raise ValueError("taddaaa")
def export_annotated_spectra_to_csv(self):
"""
id1
id2
mz
intensityy
annotation: delta mass
mz2:
decharged: True/False
similarity: intensity similaritz of mz1 vs mz2
originated mz: if decharged the original mz is saved
"""
print("start exporting information")
with open(self.output_path, "wt") as csvfile:
writr = csv.writer(csvfile, lineterminator=os.linesep)
writr.writerow(
("id1", "id2", "mz", "intensity", "annotation", "mz2",
"decharged", 'similarity', 'originated mz', "extra_mass"))
for ref in self.references:
ms = ref.create_ms(iMin_similarity=self.min_rel_similarity)
for chrg in ms.spectrum:
for key in ms.spectrum[chrg].keys():
for mp in ms.spectrum[chrg][key]:
if mp.meta['mass'] != -1:
writr.writerow(
(ref.id_1, ref.id_2, mp.mz, mp.intensity,
mp.meta['delta'], mp.meta['mass'],
mp.meta['decharged'],
mp.meta['similarity'],
mp.meta['originated_mz'], ref.extra_mass))
def export_annotated_spectra_to_mgf(self,
mgf_path,
report_just_heavy=False):
spectra_out = []
for ref in self.references:
if report_just_heavy:
ms = ref.just_create_heavy_ms()
else:
ms = ref.create_ms(iMin_similarity=self.min_rel_similarity)
buf_peaks = []
buf_int = []
for chrg in ms.spectrum:
for key in ms.spectrum[chrg].keys():
for mp in ms.spectrum[chrg][key]:
if report_just_heavy:
buf_peaks.append(mp.mz)
buf_int.append(mp.intensity)
else:
if mp.meta['mass'] != -1:
buf_peaks.append(mp.mz)
buf_int.append(mp.intensity)
if len(buf_peaks) != 0:
spectra_out.append({
'm/z array': buf_peaks,
'intensity array': buf_int,
'params': ref.params
})
mgf.write(spectra=spectra_out, output=mgf_path)
def __init__(self, path):
self.path = path
self.masterSpectrum = MasterSpectrum()
def load_msconvert_mgf(self):
"""
creates references based on precursor mass
a missing scanid means an ms1 event
by default referencing works just within one ms2 block
"""
fc = calculate_Delta_by_ppm(self.ppm)
scan_id_ary = []
problems = []
error = 0
with mgf.read(self.path) as spectra:
for spectrum in spectra:
mass = spectrum['params']['pepmass'][0]
precursor_chrg = int(spectrum['params']['charge'][0])
mass = calculatePrecursor(mass, precursor_chrg)
scanid = int(parse_scan_id(spectrum['params']['title']))
if len(scan_id_ary) == 0:
scan_id_ary.append(scanid)
else:
if scanid != scan_id_ary[-1] + 1:
if len(scan_id_ary) % 2 == 1:
problems.append(scan_id_ary[0])
error += 1
scan_id_ary = []
scan_id_ary.append(scanid)
else:
scan_id_ary = []
scan_id_ary.append(scanid)
self.ms = MasterSpectrum(
) # new MS if scan_id group (seperated by ms1) is completed
else:
scan_id_ary.append(scanid)
found = False
if len(self.ms.spectrum) == 0:
peak = Peak(mass, scanid, fc)
self.ms.add(Peak(mass,
scanid,
fc,
meta={
'ms': spectrum['m/z array'],
'int': spectrum['intensity array'],
'params': spectrum['params']
}),
charge=precursor_chrg)
found = True
else:
if (precursor_chrg in self.ms.spectrum.keys()
): # react to charge !!!!!!
if len(self.ms.spectrum[precursor_chrg]) == 0:
peak = Peak(mass, scanid, fc)
self.ms.add(Peak(mass,
scanid,
fc,
meta={
'ms': spectrum['m/z array'],
'int':
spectrum['intensity array'],
'params': spectrum['params']
}),
charge=precursor_chrg)
found = True
else:
for extra_mass in gen_allowed_mass_diff_with_sign(
n=4, z=1):
if found is False:
peak = Peak(mass + extra_mass, 0.5, fc)
if peak.key(
) in self.ms.spectrum[precursor_chrg]:
print(precursor_chrg)
idx, bin_to_ack, a, b = self.ms.binary(
peak, 0,
len(self.ms.spectrum[
precursor_chrg][peak.key()]) -
1, precursor_chrg)
if idx != -1:
self.references.add(
Reference(
ppm=self.ppm,
id_2=scanid,
id_1=self.ms.
spectrum[precursor_chrg][
peak.key()][idx].
intensity, # also scanid
peak_list_2=spectrum[
'm/z array'],
peak_list_1=self.ms.
spectrum[precursor_chrg][
peak.key()]
[idx].meta['ms'],
mass_2=mass,
mass_1=self.ms.
spectrum[precursor_chrg][
peak.key()][idx].mz,
charge=spectrum['params']
['charge'][0],
extra_mass=extra_mass,
int_list_2=spectrum[
'intensity array'],
int_list_1=self.ms.
spectrum[precursor_chrg][
peak.key()]
[idx].meta['int'],
params2=spectrum['params'],
params1=self.ms.
spectrum[precursor_chrg][
peak.key()]
[idx].meta['params']))
found = True
del (self.ms.
spectrum[precursor_chrg][
peak.key()][idx])
if len(self.ms.
spectrum[precursor_chrg][
peak.key()]) == 0:
del (self.ms.
spectrum[precursor_chrg][
peak.key()])
if len(self.ms.spectrum[
precursor_chrg]) == 0:
del (self.ms.spectrum[
precursor_chrg])
if found is False:
self.ms.add(Peak(mass,
scanid,
fc,
meta={
'ms': spectrum['m/z array'],
'int': spectrum['intensity array'],
'params': spectrum['params']
}),
charge=precursor_chrg)
if error > 0:
print(" delete valid information {0}".format(error))
def create_ms(self, iMin_similarity):
"""
int list 1 and mz list 1 are the ones with a higher precursor mass
int/mz list1 will create a master spectrum
for every peak in list2
check all allowed mass diffs (depending on charge and number of tags)
-> find peak + delta
--> for all save a similarity value (calculating ratio of int1/int2) depending which is bigger
-> sort for highest similarity score
-> check if masterspectrum peak already has an refering peak2
--> if new similarity is higher: replace
a peak must have at least int_similarity = 0.5
"""
req_min_similarity = iMin_similarity
ms = MasterSpectrum()
rel_int = calculateRelativeIntensity(self.int_list_1)
for mass, rel_int, i in zip(self.peak_list_1, rel_int,
self.int_list_1):
ms.add(
Peak(mass,
i,
self.fc,
meta={
'delta': 511,
'mass': -1,
'decharged': False,
'similarity': -1,
'originated_mz': mass,
'rel_int': rel_int
}))
num = round((abs(self.extra_mass) /
calculate_Mass_Diff())) # num of tags possible
self.params["#numtags"] = "{0}".format(num)
deltas = mass_diff_decharging_stuff(n=int(num), z=self.charge)
dd = deltas.keys()
dd = list(dd)
dd.sort()
rel_int = calculateRelativeIntensity(self.int_list_2)
for mass, rel_int, i in zip(self.peak_list_2, rel_int,
self.int_list_2):
similarity_most_similar = -1
idx_most_similar = -1
peak_key_most_similar = -1
delta_most_similar = -1
mass_most_similar = -1
for delta in dd:
peak = Peak(mass + delta, i, self.fc)
if peak.key() in ms.spectrum[0]:
idx, bin_to_ack, should_merge_left_peak, should_merge_right_peak = ms.binary(
peak, 0,
len(ms.spectrum[0][peak.key()]) - 1, 0)
if idx != -1: # found
rel_int1 = ms.spectrum[0][
peak.key()][idx].meta['rel_int']
ratio = rel_int1 / rel_int
if ratio > 1: # i want ratio to be between 0 - 1
ratio = 1 / ratio
if ratio > similarity_most_similar:
similarity_most_similar = ratio
idx_most_similar = idx
peak_key_most_similar = peak.key()
mass_most_similar = mass
delta_most_similar = delta
if similarity_most_similar > req_min_similarity:
if similarity_most_similar > ms.spectrum[0][
peak_key_most_similar][idx_most_similar].meta[
'similarity']:
ms.spectrum[0][peak_key_most_similar][
idx_most_similar].meta = {
'delta':
delta_most_similar,
'mass':
mass_most_similar,
'decharged':
deltas[delta_most_similar]['decharge']['state'],
'similarity':
similarity_most_similar,
'originated_mz':
ms.spectrum[0][peak_key_most_similar]
[idx_most_similar].mz,
'rel_int':
ratio
}
if deltas[delta_most_similar]['decharge']['state']:
ms.add(
Peak(calculatePrecursor(
ms.spectrum[0][peak_key_most_similar]
[idx_most_similar].mz,
deltas[delta_most_similar]['decharge']['z']),
ms.spectrum[0][peak_key_most_similar]
[idx_most_similar].intensity,
self.fc,
meta=ms.spectrum[0][peak_key_most_similar]
[idx_most_similar].meta))
del (ms.spectrum[0][peak_key_most_similar]
[idx_most_similar])
if len(ms.spectrum[0][peak_key_most_similar]) == 0:
del (ms.spectrum[0][peak_key_most_similar])
return ms
class MgfAnnotaterDistiller(MgfAnnotater):
"""
def __init__(self, path, output_path, ppm=10):
self.ppm = ppm
self.ms = MasterSpectrum()
self.path = path
self.out = []
self.output_path = output_path
self.references = sortedlist(key=lambda i: i.id_1)
"""
def load_improved_csv(self, path_score):
spectra_to_be_referenced = {}
with open(path_score) as csvfile:
reader = csv.DictReader(csvfile)
for row in reader:
if row['id1'] > row['id2']:
spectra_to_be_referenced[int(row['id1'])] = int(row['id2'])
else:
spectra_to_be_referenced[int(row['id2'])] = int(row['id1'])
self.ids_to_be_referenced = spectra_to_be_referenced
def load_distiller_mgf(self):
"""
creates references based on improved csv
a missing scanid means an ms1 event
"""
data = {}
alm = [i for i in gen_allowed_mass_diff_with_sign(n=4, z=1)]
with mgf.read(self.path) as spectra:
for spectrum in spectra:
mass = spectrum['params']['pepmass'][0]
precursor_chrg = int(spectrum['params']['charge'][0])
mass = calculatePrecursor(mass, precursor_chrg)
scanid = int(parse_scan_id(spectrum['params']['title']))
if scanid in self.ids_to_be_referenced:
if self.ids_to_be_referenced[scanid] in data:
mass1 = data[self.ids_to_be_referenced[scanid]]['params']['pepmass'][0]
precursor_chrg1 = int(data[self.ids_to_be_referenced[scanid]]['params']['charge'][0])
mass1 = calculatePrecursor(mass1, precursor_chrg1)
diff = abs(mass1 - mass)
diff2 = [abs(diff - abs(i)) for i in alm]
pos = diff2.index(min(diff2))
p = "mass1:\t {0}\n"
p += "mass:\t {1}\n"
p += "scanid:\t {2}\n"
p += "charge:\t {3}\n"
p += "charge2:\t {4}\n"
p += "scanid2:\t {5}\n"
if diff > 21: # distiller changes precursor charge therefore precurosr mass calculation is wrong
print(p.format(mass1, mass, scanid, precursor_chrg1, spectrum['params']['charge'][0], self.ids_to_be_referenced[scanid]))
print(diff)
print(diff2)
print("----------------")
else:
self.references.add(Reference(ppm=self.ppm,
id_2=scanid,
id_1=self.ids_to_be_referenced[scanid], # also scanid
peak_list_2=spectrum['m/z array'],
peak_list_1=data[self.ids_to_be_referenced[scanid]]['m/z array'],
mass_2=mass,
mass_1=mass1,
charge=spectrum['params']['charge'][0],
extra_mass=alm[pos],
int_list_2=spectrum['intensity array'],
int_list_1=data[self.ids_to_be_referenced[scanid]]['intensity array'],
params2=spectrum['params'],
params1=data[self.ids_to_be_referenced[scanid]]['params']))
del(data[self.ids_to_be_referenced[scanid]])
del(self.ids_to_be_referenced[scanid])
else:
data[scanid] = spectrum
def load_distiller_mgf2(self):
"""
creates references based on precursor mass
a missing scanid means an ms1 event
by default referencing works just within one ms2 block
"""
fc = calculate_Delta_by_ppm(self.ppm)
error = 0
self.ms = MasterSpectrum()
with mgf.read(self.path) as spectra:
for spectrum in spectra:
mass = spectrum['params']['pepmass'][0]
precursor_chrg = int(spectrum['params']['charge'][0])
mass = calculatePrecursor(mass, precursor_chrg)
scanid = int(parse_scan_id(spectrum['params']['title']))
found = False
if len(self.ms.spectrum) == 0:
peak = Peak(mass, scanid, fc)
self.ms.add(Peak(mass, scanid, fc, meta={'ms': spectrum['m/z array'], 'int': spectrum['intensity array'], 'params': spectrum['params']}),
charge=precursor_chrg)
found = True
else:
if (precursor_chrg in self.ms.spectrum.keys()): # react to charge !!!!!!
if len(self.ms.spectrum[precursor_chrg]) == 0:
peak = Peak(mass, scanid, fc)
self.ms.add(Peak(mass, scanid, fc, meta={'ms': spectrum['m/z array'], 'int': spectrum['intensity array'], 'params': spectrum['params']}),
charge=precursor_chrg)
found = True
else:
for extra_mass in gen_allowed_mass_diff_with_sign(n=4, z=1):
if found is False:
peak = Peak(mass + extra_mass, 0.5, fc)
if peak.key() in self.ms.spectrum[precursor_chrg]:
print(precursor_chrg)
idx, bin_to_ack, a, b = self.ms.binary(peak, 0, len(self.ms.spectrum[precursor_chrg][peak.key()]) - 1, precursor_chrg)
if idx != -1:
self.references.add(Reference(ppm=self.ppm,
id_2=scanid,
id_1=self.ms.spectrum[precursor_chrg][peak.key()][idx].intensity, # also scanid
peak_list_2=spectrum['m/z array'],
peak_list_1=self.ms.spectrum[precursor_chrg][peak.key()][idx].meta['ms'],
mass_2=mass,
mass_1=self.ms.spectrum[precursor_chrg][peak.key()][idx].mz,
charge=spectrum['params']['charge'][0],
extra_mass=extra_mass,
int_list_2=spectrum['intensity array'],
int_list_1=self.ms.spectrum[precursor_chrg][peak.key()][idx].meta['int'],
params2=spectrum['params'],
params1=self.ms.spectrum[precursor_chrg][peak.key()][idx].meta['params']))
found = True
del(self.ms.spectrum[precursor_chrg][peak.key()][idx])
if len(self.ms.spectrum[precursor_chrg][peak.key()]) == 0:
del(self.ms.spectrum[precursor_chrg][peak.key()])
if len(self.ms.spectrum[precursor_chrg]) == 0:
del(self.ms.spectrum[precursor_chrg])
if found is False:
limit_scan_id = scanid - 20 # could start at -19
ms_bac = MasterSpectrum()
for chrg in self.ms.spectrum:
for key in self.ms.spectrum[chrg].keys():
for mp in self.ms.spectrum[chrg][key]:
if mp.intensity >= limit_scan_id:
ms_bac.add(mp, charge=chrg)
self.ms = ms_bac
self.ms.add(Peak(mass, scanid, fc, meta={'ms': spectrum['m/z array'], 'int': spectrum['intensity array'], 'params': spectrum['params']}),
charge=precursor_chrg)
if error > 0:
print(" delete valid information {0}".format(error))
