def testSplitBy():
t = ms.toTable("a", [1, 1, 3, 4])
t.addColumn("b", [1, 1, 3, 3])
t.addColumn("c", [1, 2, 1, 4])
t._print()
subts = t.splitBy("a")
assert len(subts) == 3
res = ms.mergeTables(subts)
assert len(res) == len(t)
subts[0]._print()
assert res.a.values == t.a.values
assert res.b.values == t.b.values
assert res.c.values == t.c.values
# check if input tables are not altered
for subt in subts:
assert subt.getColNames() == ["a", "b", "c"]
subts = t.splitBy("a", "c")
assert len(subts) == 4
res = ms.mergeTables(subts)
assert res.a.values == t.a.values
assert res.b.values == t.b.values
assert res.c.values == t.c.values
# check if input tables are not altered
for subt in subts:
assert subt.getColNames() == ["a", "b", "c"]
def testSplitBy():
t = ms.toTable("a", [1,1,3,4])
t.addColumn("b", [1,1,3,3])
t.addColumn("c", [1,2,1,4])
t._print()
subts = t.splitBy("a")
assert len(subts) == 3
res = ms.mergeTables(subts)
assert len(res) == len(t)
subts[0]._print()
assert res.a.values == t.a.values
assert res.b.values == t.b.values
assert res.c.values == t.c.values
# check if input tables are not altered
for subt in subts:
assert subt.getColNames() == [ "a", "b", "c"]
subts = t.splitBy("a", "c")
assert len(subts) == 4
res = ms.mergeTables(subts)
assert res.a.values == t.a.values
assert res.b.values == t.b.values
assert res.c.values == t.c.values
# check if input tables are not altered
for subt in subts:
assert subt.getColNames() == [ "a", "b", "c"]
def integrate_matching_fragments(tables, transitions, delta_mz):
"""extracts mass transitions defined in transition table 'transitions'
with column names 'name', 'precursor', 'fragment', 'rtmin', 'rtmax'.
Detecting the peaks does not use a peak detector as centWave, but uses mz
and rt ranges to fit a EMG model to the underlying raw peaks m/z traces.
This avoids cumbersome parameter optimization of a peak detector and
returns all peaks irrespective of filtering according to some heuristic
criterion for peak quality.
"""
table = ms.mergeTables(tables)
# To shorten the name in commands we introduce abbreviations
trans = transitions
t = table
# table has columns
#
# precursor fragment peakmap mzmin mzmax
#
# and transitions:
#
# name precursor fragment rtmin rtmax
# find candidates defined in parameter table transitions
identified = t.join(trans,
trans.precursor.approxEqual(t.precursor, delta_mz)\
& trans.fragment.approxEqual(t.fragment, delta_mz))
# identified has columns:
# precursor fragment peakman mzmin mzmax name__0 precursor__0
# ... fragment__0 rtmin__0 rtmax__0
identified.renameColumns(name__0="name",
rtmin__0="rtmin",
rtmax__0="rtmax")
identified = identified.extractColumns("name", "precursor", "fragment",
"mzmin", "mzmax", "rtmin", "rtmax",
"peakmap")
# now we have columns mzmin, mzmax, rtmin, rtmax and peakmap which we
# need for fitting EMG model with ms.integrate:
identified = ms.integrate(identified, "emg_exact", msLevel=2)
return identified.splitBy("name")
def process_srm_data(peakmap, n_digits):
tables = split_srm_peakmap_to_tables(peakmap, n_digits)
# Each table in tables now contains chromatograms for each precursor
# We flatten each of these tables to one row tables, so that
# the table explorer overlays these chromatograms if you select
# a specific row.
flattened_tables = [flatten(t) for t in tables]
# Concatenate all these one row table to one big result table:
merged = ms.mergeTables(flattened_tables)
# For later visual inspection with the table explorer a title is helpful.
# Here we use the name of the peakmap file:
merged.title = "SRM/MRM analysis of %s" % peakmap.meta["source"]
return merged
def process_srm_data(peakmap, n_digits):
tables = split_srm_peakmap_to_tables(peakmap, n_digits)
# Each table in tables now contains chromatograms for each precursor
# We flatten each of these tables to one row tables, so that
# the table explorer overlays these chromatograms if you select
# a specific row.
flattened_tables = [flatten(t) for t in tables]
# Concatenate all these one row table to one big result table:
merged = ms.mergeTables(flattened_tables)
# For later visual inspection with the table explorer a title is helpful.
# Here we use the name of the peakmap file:
merged.title = "SRM/MRM analysis of %s" % peakmap.meta["source"]
return merged
def integrate_matching_fragments(tables, transitions, delta_mz):
"""extracts mass transitions defined in transition table 'transitions'
with column names 'name', 'precursor', 'fragment', 'rtmin', 'rtmax'.
Detecting the peaks does not use a peak detector as centWave, but uses mz
and rt ranges to fit a EMG model to the underlying raw peaks m/z traces.
This avoids cumbersome parameter optimization of a peak detector and
returns all peaks irrespective of filtering according to some heuristic
criterion for peak quality.
"""
table = ms.mergeTables(tables)
# To shorten the name in commands we introduce abbreviations
trans = transitions
t = table
# table has columns
#
# precursor fragment peakmap mzmin mzmax
#
# and transitions:
#
# name precursor fragment rtmin rtmax
# find candidates defined in parameter table transitions
identified = t.join(trans,
trans.precursor.approxEqual(t.precursor, delta_mz)\
& trans.fragment.approxEqual(t.fragment, delta_mz))
# identified has columns:
# precursor fragment peakman mzmin mzmax name__0 precursor__0
# ... fragment__0 rtmin__0 rtmax__0
identified.renameColumns(name__0="name",
rtmin__0="rtmin",
rtmax__0="rtmax")
identified = identified.extractColumns("name", "precursor", "fragment",
"mzmin", "mzmax", "rtmin", "rtmax",
"peakmap")
# now we have columns mzmin, mzmax, rtmin, rtmax and peakmap which we
# need for fitting EMG model with ms.integrate:
identified = ms.integrate(identified, "emg_exact", msLevel=2)
return identified.splitBy("name")
def testMerge():
t1 = ms.toTable("a", [1])
t1.addColumn("b", [2])
t1.addColumn("c", [3])
t2 = ms.toTable("a", [1,2])
t2.addColumn("c", [1,3])
t2.addColumn("d", [1,4])
tn = ms.mergeTables([t1, t2])
assert tn.a.values == [1, 1, 2]
assert tn.b.values == [2, None, None]
assert tn.c.values == [3, 1, 3]
assert tn.d.values == [None, 1, 4]
# check if input tables are not altered
assert t1.getColNames() == [ "a", "b", "c"]
assert t2.getColNames() == [ "a", "c", "d"]
def testMerge():
t1 = ms.toTable("a", [1])
t1.addColumn("b", [2])
t1.addColumn("c", [3])
t2 = ms.toTable("a", [1, 2])
t2.addColumn("c", [1, 3])
t2.addColumn("d", [1, 4])
tn = ms.mergeTables([t1, t2])
assert tn.a.values == [1, 1, 2]
assert tn.b.values == [2, None, None]
assert tn.c.values == [3, 1, 3]
assert tn.d.values == [None, 1, 4]
# check if input tables are not altered
assert t1.getColNames() == ["a", "b", "c"]
assert t2.getColNames() == ["a", "c", "d"]
def process_srm_data(peakmap, transitions, delta_mz):
tables = split_srm_peakmap_by_precursors_to_tables(peakmap, delta_mz)
# Each table in tables now contains chromatograms for each precursor
# We flatten each of these tables to one row tables, so that
# the table explorer overlays these chromatograms if you select
# a specific row.
if transitions is not None:
tables = integrate_matching_fragments(tables, transitions, delta_mz)
else:
tables = integrate_fragments(tables)
flattened_tables = [flatten(t) for t in tables]
# Concatenate all these one row table to one big result table:
merged = ms.mergeTables(flattened_tables)
# For later visual inspection with the table explorer a title is helpful.
# Here we use the name of the peakmap file:
merged.title = "SRM peaks of %s" % peakmap.meta["source"]
return merged
def process_srm_data(peakmap, transitions, delta_mz):
tables = split_srm_peakmap_by_precursors_to_tables(peakmap, delta_mz)
# Each table in tables now contains chromatograms for each precursor
# We flatten each of these tables to one row tables, so that
# the table explorer overlays these chromatograms if you select
# a specific row.
if transitions is not None:
tables = integrate_matching_fragments(tables, transitions, delta_mz)
else:
tables = integrate_fragments(tables)
flattened_tables = [flatten(t) for t in tables]
# Concatenate all these one row table to one big result table:
merged = ms.mergeTables(flattened_tables)
# For later visual inspection with the table explorer a title is helpful.
# Here we use the name of the peakmap file:
merged.title = "SRM peaks of %s" % peakmap.meta["source"]
return merged
