def test_GaussNoiseInjector_generate_intensity_noise():
i_array = np.array([1e6, 2e6, 3e6], dtype="float32")
scan = Scan({
"mz": np.array([100, 200, 300], dtype="float32"),
"i": i_array,
"ms_level": 1,
})
noise_injector = GaussNoiseInjector()
noise = noise_injector._generate_intensity_noise(scan)
assert (abs(noise) < i_array * 0.2).all()
def test_GaussNoiseInjector_intensity_noise_ms2():
i_array = np.array([1e6, 2e6, 3e6], dtype="float32")
scan = Scan({
"mz": np.array([100, 200, 300], dtype="float32"),
"i": i_array,
"ms_level": 2,
})
noise_injector = GaussNoiseInjector()
scan = noise_injector.inject_noise(scan, dropout=0.0)
assert (scan.i - i_array < i_array * 0.2).all()
def test_GaussNoiseInjector_generate_mz_noise():
i_array = np.array([1e6, 2e6, 3e6], dtype="float32")
scan = Scan({
"mz": np.array([100, 200, 300], dtype="float32"),
"i": i_array,
"ms_level": 1,
})
noise_injector = GaussNoiseInjector()
noise = noise_injector._generate_mz_noise(scan)
# breakpoint()
assert (abs(noise) < scan.mz * 1e-5).all()
def test_GaussNoiseInjector_intensity_noise_ms2_drop_all():
i_array = np.array([1e6, 2e6, 3e6], dtype="float32")
scan = Scan({
"mz": np.array([100, 200, 300], dtype="float32"),
"i": i_array,
"ms_level": 2,
})
noise_injector = GaussNoiseInjector()
scan = noise_injector.inject_noise(scan, dropout=1.0)
assert len(scan.mz) == 0
assert len(scan.i) == 0
def test_gauss_mz_noise_ms1():
i_array = np.array([1e6, 2e6, 3e6], dtype="float32")
scan = Scan({
"mz": np.array([100, 200, 300], dtype="float32"),
"i": i_array,
"ms_level": 1,
})
noise_injector = GaussNoiseInjector()
scan = noise_injector.inject_noise(scan)
# breakpoint()
assert (abs(scan.mz - np.array([100, 200, 300], dtype="float32")) <
np.array([100, 200, 300], dtype="float32") * 1e-5).all()
def test_GaussNoiseInjector_intensity_noise_ms1():
i_array = np.array([1e6, 2e6, 3e6], dtype="float32")
scan = Scan({
"mz": np.array([100, 200, 300], dtype="float32"),
"i": i_array,
"ms_level": 1,
})
noise_injector = GaussNoiseInjector(variance=0.2)
scan = noise_injector.inject_noise(scan)
# breakpoint()
# assert abs((scan.i - 3e6) < 3e6 * 0.05).all()
assert (scan.i - i_array < i_array * 0.2).all()
def test_write_mzml_get_TIC():
file = NamedTemporaryFile("wb")
peak_props = {
"inosine": {
"chemical_formula": "+C(10)H(12)N(4)O(5)",
"trivial_name": "inosine",
"charge": 2,
"scan_start_time": 0,
"peak_width": 30, # seconds
"peak_function": None,
"peak_params": {},
}
}
mzml_params = {
"gradient_length": 30,
}
noise_injector = GaussNoiseInjector(variance=0.0)
mzml_path = write_mzml(file, peak_props, fragmentor, noise_injector,
mzml_params)
reader = pymzml.run.Reader(mzml_path)
assert reader.get_spectrum_count() == 1000
reader = pymzml.run.Reader(mzml_path)
tics = []
for spec in reader:
if spec.ms_level == 1:
tics.append(sum(spec.i))
tic = reader["TIC"]
assert tic.peaks()[:, 1] == pytest.approx(np.array(tics, dtype="float32"))
def main():
file = NamedTemporaryFile("wb")
peak_props = {
"ELVISLIVES": {
"trivial_name": "ELVISLIVES",
"charge": 2,
"scan_start_time": 0,
"peak_width": 30, # seconds
"peak_function": "gauss",
"peak_params": {
"sigma": 3
}, # 10% of peak width
},
"ELVISLIVSE": {
"charge": 2,
"trivial_name": "ELVISLIVSE",
"scan_start_time": 15,
"peak_width": 30, # seconds
"peak_function": "gauss",
"peak_params": {
"sigma": 3
}, # 10% of peak width,
},
}
mzml_params = {
"gradient_length": 45,
}
fragmentor = PeptideFragmentor()
noise_injector = GaussNoiseInjector(variance=0.05)
mzml_path = write_mzml(file, peak_props, fragmentor, noise_injector,
mzml_params)
def test_write_2_mols_same_cc():
file = NamedTemporaryFile("wb")
peak_props = {
"uridine": {
"charge": 2,
"chemical_formula": "+C(9)H(11)N(2)O(6)",
"trivial_name": "uridine",
"scan_start_time": 0,
"peak_width": 30, # seconds
"peak_function": "gauss",
"peak_params": {
"sigma": 1
}, # 10% of peak width,
"peak_scaling_factor": 0.5 * 1e6,
},
"pseudouridine": {
"chemical_formula": "+C(9)H(11)N(2)O(6)",
"trivial_name": "pseudouridine",
"charge": 2,
"scan_start_time": 15,
"peak_width": 30, # seconds
"peak_function": "gauss",
"peak_params": {
"sigma": 1
}, # 10% of peak width,
"peak_scaling_factor": 1e6,
},
}
mzml_params = {
"gradient_length": 45,
"min_intensity": 10,
}
mzml_params = {
"gradient_length": 45,
}
fragmentor = NucleosideFragmentor()
noise_injector = GaussNoiseInjector(variance=0.0)
mzml_path = write_mzml(file, peak_props, fragmentor, noise_injector,
mzml_params)
reader = pymzml.run.Reader(mzml_path)
# assert reader.get_spectrum_count() == 1499
intensities = []
for spec in reader:
if spec.ms_level == 1:
if len(spec.i) > 0:
intensities.append(sum(spec.i))
peaks, _ = find_peaks(intensities)
assert len(peaks) == 2
def test_write_inosine_adenosine_gauss_shift_mzml():
file = NamedTemporaryFile("wb")
peak_props = {
"inosine": {
"chemical_formula": "+C(10)H(12)N(4)O(5)",
"charge": 1,
"trivial_name": "inosine",
"scan_start_time": 0,
"peak_width": 30, # seconds
"peak_function": "gauss",
"peak_params": {
"sigma": 3
}, # 10% of peak width,
"peak_scaling_factor": 1e6,
},
"+C(10)H(13)N(5)O(4)": {
"chemical_formula": "+C(10)H(13)N(5)O(4)",
"trivial_name": "adenosine",
"charge": 1,
"scan_start_time": 15,
"peak_width": 30, # seconds
"peak_function": "gauss",
"peak_params": {
"sigma": 3
}, # 10% of peak width,
"peak_scaling_factor": 1e6,
},
}
mzml_params = {
"gradient_length": 45,
}
noise_injector = GaussNoiseInjector(variance=0.0)
# noise_injector = UniformformNoiseInjector()
mzml_path = write_mzml(file, peak_props, fragmentor, noise_injector,
mzml_params)
reader = pymzml.run.Reader(mzml_path)
assert reader.get_spectrum_count() == 1500
ino_rt = []
ino_intensities = []
adeno_rt = []
adeno_intensities = []
ino_mono = 269.0880
adeno_mono = 268.1040
for spec in reader:
if spec.ms_level == 1:
ino_i = spec.i[abs(spec.mz - ino_mono) < 0.001]
adeno_i = spec.i[abs(spec.mz - adeno_mono) < 0.001]
if len(adeno_i) > 0:
adeno_intensities.append(adeno_i[0])
adeno_rt.append(spec.scan_time[0])
if len(ino_i) > 0:
ino_intensities.append(ino_i[0])
ino_rt.append(spec.scan_time[0])
# assert rt max diff is about 15
m_i = np.argmax(ino_intensities)
m_a = np.argmax(adeno_intensities)
mean_i_rt = ino_rt[m_i]
mean_a_rt = adeno_rt[m_a]
assert 14 < (mean_a_rt - mean_i_rt) < 16
generate_molecule_isotopologue_lib,
generate_scans,
write_mzml,
)
class TestFragmentor(AbstractFragmentor):
def __init__(self):
print("Fragmentor goes chop chop chop chop chop ...")
def fragment(self, mol):
return np.array([(200, 1e5)])
fragmentor = TestFragmentor()
noise_injector = GaussNoiseInjector(variance=0.05)
def test_write_mzml():
"""Write a mzML without spectra readable by pymzML."""
file = NamedTemporaryFile("wb")
peak_props = {}
mzml_params = {
"gradient_length": 0,
}
mzml_path = write_mzml(file, peak_props, fragmentor, noise_injector,
mzml_params)
reader = pymzml.run.Reader(mzml_path)
assert reader.get_spectrum_count() == 0