def parse_mz_id(self):
data = mzid.read(self.path)
max_rank = 2
for d in data:
title = d['spectrum title']
scan_id = parse_scan_id(title)
fragments = []
len_frag = len(d['SpectrumIdentificationItem'])
pos = 1
while(pos <= min(max_rank, len_frag)):
for fragmentation in d['SpectrumIdentificationItem'][pos - 1]['IonType']: # 0 because just first rank
for f in fragmentation['FragmentArray']:
if f['measure_ref'] == 'm_mz':
mz = f['values']
elif f['measure_ref'] == 'm_error':
error = f['values']
else:
pass
fragments.append(Fragment(name=fragmentation['name'],
indice=fragmentation['index'],
charge=fragmentation['charge'],
mz=mz,
error=error,
scanid=parse_scan_id(title)))
if scan_id in self.scan1.data:
self.scan1.data[scan_id][pos] = fragments
else:
self.scan1.data[scan_id] = {}
self.scan1.data[scan_id][pos] = fragments
pos += 1
def parse_mz_id(self):
"""
reading mzid
saving every spectrum identification (but just rank 1)
returns:
None
"""
data = mzid.read(self.path)
for d in data:
title = parse_scan_id(d['spectrum title'])
ident = {}
len_ranks = len(d['SpectrumIdentificationItem'])
if len_ranks > 1:
for i in [0, 1]:
identification = d['SpectrumIdentificationItem'][i] # 0 because just first rank
peptide_ref = identification['peptide_ref']
ident[i + 1] = Identification(mzid_info_lvl_fragmentation=identification['IonType'],
peptide_ref=peptide_ref,
title=title)
else:
identification = d['SpectrumIdentificationItem'][0] # 0 because just first rank
peptide_ref = identification['peptide_ref']
ident[1] = Identification(mzid_info_lvl_fragmentation=identification['IonType'],
peptide_ref=peptide_ref,
title=title)
self.identifications.append(ident)
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 read_enhanced_spectrum(self, path):
"""
saving a masterSpectrum for every spectrum is not a memory efficient idea
4 GB for 8247 spectra
instead creating on request (saving reference and spectra object)
"""
with mgf.read(path) as spectra:
for spectrum in spectra:
# charge_of_spectrum = str(spectrum['params']['charge'][0])
scan_id = parse_scan_id(spectrum['params']['title'])
self.mgf_reads[scan_id] = Reference(scan_id, spectrum['m/z array'])
def select_mgf(self):
spectra_out = []
with mgf.read(self.path_mgf_in) as spectra:
for spectrum in spectra:
scanid = int(parse_scan_id(spectrum['params']['title']))
if scanid in self.list_chosen:
spectra_out.append({'m/z array': spectrum['m/z array'],
'intensity array': spectrum['intensity array'],
'params': spectrum['params']})
mgf.write(spectra=spectra_out, output=self.path_mgf_out)
def load_export(self):
with mgf.read(self.mgf_path) as spectra, open(self.output_csv,
"wt") as csvfile:
writr = csv.writer(csvfile, lineterminator=os.linesep)
writr.writerow(("scanid", "peak", "rel_int"))
for spectrum in spectra:
scan_id = parse_scan_id(spectrum['params']['title'])
st = "scanid:\t{0}\n"
print(st.format(scan_id))
rel_int = calculateRelativeIntensity(
spectrum['intensity array'])
for m, i in zip(spectrum['m/z array'], rel_int):
writr.writerow((scan_id, m, i))
def test_splitting(self):
test = "\"controllerType=0 controllerNumber=1 scan=1316\""
assert_equal(parse_scan_id(test), '1316')
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 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))