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 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 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 analyse_mzid_vs_mgf(self):
delta_func = calculate_Delta_by_ppm(20)
if self.mgf_reads == {}:
raise ValueError("need the mgf beforehand read_enhanced_spectrum")
else:
for ids in self.identifications:
mzs = ids.report_all_mzs()
ms = self.mgf_reads[ids.scan_id].request_ms()
for mz in mzs:
peak = Peak(mz, 2, delta_func)
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:
error = "mz:\t{0}\nscan_id:\t{1}".format(mz, ids.scan_id)
raise ValueError(error)
else:
pass
else:
error = "mz:\t{0}\nscan_id:\t{1}".format(mz, ids.scan_id)
raise ValueError(error)
def analyze_mzid_id_vs_score_file(self, path_score, output_path="/home/tobiass/Desktop/out.csv"):
"""
Prerequirenment:
loading mzid(spectrum and peptide info)
input:
path to score file (created by control.py improve_csv)
Description:
- finding all important scan ids
- loading all scans as spectra from score file
- saving delta as meta information for every peak
-
cheching all identifiacation (--> peptide_ref)
- for peptide ref:
read series (b/y, index and charge, mz)
generate tag amount for every position
tag series of b/y with tmt diff
for series:
check in scans for delta
Output:
none
"""
with open(output_path, "wt") as csvfile:
writr = csv.writer(csvfile, lineterminator=os.linesep)
writr.writerow(("scanid", "rank", "peak", "position", "frag", "expected by mascot", "found", "charge", "PeaksMatchedInSerie", "max_number_of_tmt_tag"))
valid_scan_ids = {}
delta_func = calculate_Delta_by_ppm(20)
for identification_hashobject in self.identifications:
valid_scan_ids[identification_hashobject[1].scan_id] = True
valid_scan_ids = valid_scan_ids.keys() # quick and dirty - smarter with some kind of unique tree
valid_scan_ids = [int(i) for i in valid_scan_ids]
valid_spectra_info = self.read_score_file(path_score, valid_scan_ids)
self.parse_mz_id_peptide_ref()
for identification_hashobject in self.identifications:
for i in identification_hashobject:
peptide_ref = identification_hashobject[i].peptide_ref
scan_id = int(identification_hashobject[i].scan_id)
b_tmt, y_tmt = self.peptide_evidence[peptide_ref].get_annotated_positions()
max_tmt_tag = calculate_max_tmt(b_tmt) # same num for b and y
ms, dPeaks_matched = generateMS_by_score_file(valid_spectra_info[scan_id])
for ion_serie in identification_hashobject[i].ion_series_ary:
tmt_masses = calculate_allowed_Mass_Diff(n=max_tmt_tag, z=ion_serie.charge)
if 'b' in ion_serie.fragtype:
tmt_pos_ary = b_tmt
elif 'y' in ion_serie.fragtype:
tmt_pos_ary = y_tmt
else:
raise ValueError("{0}\tis not valid fragtype".format(ion_serie.fragtype))
for pos, mz in zip(ion_serie.ions_index, ion_serie.mz_ary):
num_tag_at_pos = tmt_pos_ary[pos]
if num_tag_at_pos == 0:
expected_mass_delta = 0
else:
expected_mass_delta = tmt_masses[num_tag_at_pos][ion_serie.charge]
peak = Peak(mz, 1, delta_func)
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:
# just heavy spectra are just all peaks of the heavy labeled partner of a mix (there is no garantee that both
# have the same number of peaks
# BUT the number of spectra is the same
msg = "Peak:\t{0}\n"
msg += "frag:\t{4}\n"
msg += "position:\t{5}\n"
msg += "scanid:\t{1}\n"
msg += "expected:\t{2}\n"
msg += "found:\t{3}\n"
msg += "peak was just part of heavy spectra(but same number before \".\"\n"
# print(msg.format(mz, scan_id, expected_mass_delta, -2,
# ion_serie.fragtype, pos))
writr.writerow((scan_id, i, mz, pos, ion_serie.fragtype, expected_mass_delta, -2, ion_serie.charge, dPeaks_matched, max_tmt_tag))
# raise ValueError("identified Peak:\t{0}\nscanid:\t{1}\ncouldnt be found in spectra".format(mz, scan_id, expected_mass_delta, -1,
# ion_serie.fragtype, pos))
else:
found_mass_diff = ms.spectrum[0][peak.key()][idx].meta
msg = "Peak:\t{0}\n"
msg += "frag:\t{4}\n"
msg += "position:\t{5}\n"
msg += "scanid:\t{1}\n"
msg += "expected:\t{2}\n"
msg += "found:\t{3}\n"
if found_mass_diff != expected_mass_delta:
writr.writerow((scan_id, i, mz, pos, ion_serie.fragtype, expected_mass_delta, found_mass_diff, ion_serie.charge, dPeaks_matched, max_tmt_tag))
# raise ValueError(msg.format(mz, scan_id, expected_mass_delta, found_mass_diff, ion_serie.fragtype, pos))
else:
# here is my goal
writr.writerow((scan_id, i, mz, pos, ion_serie.fragtype, expected_mass_delta, found_mass_diff, ion_serie.charge, dPeaks_matched, max_tmt_tag))
else:
# just heavy spectra are just all peaks of the heavy labeled partner of a mix (there is no garantee that both
# have the same number of peaks
# BUT the number of spectra is the same
msg = "Peak:\t{0}\n"
msg += "frag:\t{4}\n"
msg += "position:\t{5}\n"
msg += "scanid:\t{1}\n"
msg += "expected:\t{2}\n"
msg += "found:\t{3}\n"
msg += "peak was just part of heavy spectra\n"
# print(msg.format(mz, scan_id, expected_mass_delta, -1,
# ion_serie.fragtype, pos))
writr.writerow((scan_id, i, mz, pos, ion_serie.fragtype, expected_mass_delta, -1, ion_serie.charge, dPeaks_matched, max_tmt_tag))
def test_key(self):
p = Peak(100, 0.5, calculate_Delta_based_MZ)
assert_not_equal(p.key(), 99)
assert_equal(p.key(), 100)
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
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))