def test_runuis_nonswath_rangetree(self):
self.assertEqual(len(self.precursors_to_evaluate), 905)
par = self.par
cursor = self.db.cursor()
# If we dont use the DB, we use the rangetree to query and get our list of
# precursors that are interfering. In SWATH we dont include a +/- q1_window
# around our range or precursors because the precursor window is fixed to
# (min_q1,max_q1) and no other precursors are considered.
self.myprecursors.getFromDB(par, cursor, self.min_q1 - par.q1_window, self.max_q1 + par.q1_window)
rtree = self.myprecursors.build_rangetree()
prepare = []
for precursor in self.precursors_to_evaluate:
q3_low, q3_high = par.get_q3range_transitions()
transitions = precursor.calculate_transitions(q3_low, q3_high)
nr_transitions = len(transitions)
# Use the rangetree, whether it is swath or not
collisions_per_peptide = self.myprecursors.get_collisions_per_peptide_from_rangetree(precursor, precursor.q1 - par.q1_window, precursor.q1 + par.q1_window, transitions, par, rtree)
non_uis_list = collider.get_nonuis_list(collisions_per_peptide, par.max_uis)
for order in range(1,min(par.max_uis+1, nr_transitions+1)):
prepare.append( (len(non_uis_list[order]), collider.choose(nr_transitions,
order), precursor.parent_id , order, -1) )
self.assertEqual(len(prepare), 905*par.max_uis)
self.assertEqual(prepare[0], (0, 17.0, 1, 1, -1))
final_report = self.get_final_report(par, prepare)
self.check_final_report_nonswath(final_report)
def test_runuis_nonswath(self):
self.assertEqual(len(self.precursors_to_evaluate), 905)
par = self.par
cursor = self.db.cursor()
prepare = []
for precursor in self.precursors_to_evaluate:
q3_low, q3_high = par.get_q3range_transitions()
transitions = precursor.calculate_transitions(q3_low, q3_high)
nr_transitions = len(transitions)
precursors_obj = self.acollider._get_all_precursors(par, precursor, cursor)
collisions_per_peptide = collider.get_coll_per_peptide_from_precursors(self.acollider,
transitions, precursors_obj, par, precursor)
non_uis_list = collider.get_nonuis_list(collisions_per_peptide, par.max_uis)
for order in range(1,min(par.max_uis+1, nr_transitions+1)):
prepare.append( (len(non_uis_list[order]), collider.choose(nr_transitions,
order), precursor.parent_id , order, -1) )
self.assertEqual(len(prepare), 905*par.max_uis)
self.assertEqual(prepare[0], (0, 17.0, 1, 1, -1))
final_report = self.get_final_report(par, prepare)
self.check_final_report_nonswath(final_report)
def test_runuis_swath(self):
self.assertEqual(len(self.precursors_to_evaluate), 905)
swath_mode = False
par = self.par
R = self.R
cursor = self.db.cursor()
prepare = []
self.min_q1 = 500
self.max_q1 = 525
# Get the precursors (now for 500-525 instead of the full range)
###########################################################################
myprecursors = Precursors()
cursor = self.db.cursor()
myprecursors.getFromDB(par, cursor, self.min_q1 - par.q1_window, self.max_q1 + par.q1_window)
rtree = myprecursors.build_rangetree()
self.precursors_to_evaluate = myprecursors.getPrecursorsToEvaluate(self.min_q1, self.max_q1)
self.assertEqual(len(self.precursors_to_evaluate), 39)
isotope_correction = par.isotopes_up_to * R.mass_diffC13 / min(par.parent_charges)
temp_precursors = Precursors()
temp_precursors.getFromDB(par, self.db.cursor(), self.min_q1 - isotope_correction, self.max_q1)
all_swath_precursors = []
for p in temp_precursors.precursors:
if(p.included_in_isotopic_range(self.min_q1, self.max_q1, par) ):
all_swath_precursors.append(p)
for precursor in self.precursors_to_evaluate:
q3_low, q3_high = par.get_q3range_transitions()
transitions = precursor.calculate_transitions(q3_low, q3_high)
nr_transitions = len(transitions)
if par.ssrcalc_window > 1000:
precursors_obj = [p for p in all_swath_precursors if p.transition_group != precursor.transition_group]
else:
ssrcalc_low = precursor.ssrcalc - par.ssrcalc_window
ssrcalc_high = precursor.ssrcalc + par.ssrcalc_window
precursors_obj = [p for p in all_swath_precursors if p.transition_group != precursor.transition_group
and p.ssrcalc > ssrcalc_low and p.ssrcalc < ssrcalc_high ]
collisions_per_peptide = collider.get_coll_per_peptide_from_precursors(self.acollider,
transitions, precursors_obj, par, precursor)
non_uis_list = collider.get_nonuis_list(collisions_per_peptide, par.max_uis)
for order in range(1,min(par.max_uis+1, nr_transitions+1)):
prepare.append( (len(non_uis_list[order]), collider.choose(nr_transitions,
order), precursor.parent_id , order, -1) )
self.assertEqual(len(prepare), 39*par.max_uis)
self.assertEqual(prepare[0], (5, 8.0, 69, 1, -1) )
final_report = self.get_final_report(par, prepare)
self.check_final_report_swath(final_report)
def test_choose(self):
self.assertEqual(10, collider.choose( 5,2) )
self.assertEqual(10, collider.choose( 5,3) )
self.assertEqual(45, collider.choose(10,2) )
self.assertEqual(120, collider.choose(10,3) )
self.assertEqual(210, collider.choose(10,4) )
elif not use_experimental_height:
# We dont have experimental height data and cannot use C++ code
collisions_per_peptide = collider.get_coll_per_peptide(mycollider,
transitions, par, peptide_obj, cursor)
min_needed = mycollider._sub_getMinNeededTransitions(par, transitions, collisions_per_peptide)
#min_needed = mycollider.getMinNeededTransitions_direct(par, transitions, precursors)
else:
# here we consider the case that we have measured a number of
# transitions experimentally and want to know how many of them are
# sufficient to establish uniqueness. For this, all we need is
# that one tuple of transitions establishes uniqueness since we
# were able to measure it above the background noise.
collisions_per_peptide = collider.get_coll_per_peptide(mycollider,
transitions, par, pep, cursor)
for order in range(1,nr_transitions+1):
mymax = collider.choose(nr_transitions, order)
if use_cpp: non_uis = c_getnonuis.get_non_uis(collisions_per_peptide, order)
else:
non_uis = set()
for pepc in collisions_per_peptide.values():
get_non_uis(pepc, non_uis, i)
if len(non_uis) < mymax: break
if len(non_uis) < mymax: min_needed = order
else: min_needed = -1
spectrum.score = min_needed * nr_transitions
spectrum.min_needed = min_needed
if min_needed != -1: spectrum.score = nr_transitions - min_needed
if not par.quiet: progressm.update(1)
get_min_tr_time += time.time() - tmp_time; tmp_time = time.time()
"qadd": qadd,
}
print q
cursor.execute(q)
print "Obtained all peptides"
alltuples = list(cursor.fetchall())
progressm = progress.ProgressMeter(total=len(alltuples), unit="peptides")
prepare = []
total_transitions = 0
total_assays = 0
for kk, pep in enumerate(alltuples):
q3_low, q3_high = [400, 1200]
q3_low, q3_high = [400, 1400]
q3charges = [1, 2]
q3charges = [1]
# q3_low, q3_high = [0, 12000]
#
# new way to calculate the precursors
transitions = c_getnonuis.calculate_transitions_ch(((-2, pep[0], -1),), q3charges, q3_low, q3_high)
# print pep, len(pep[0]), len(transitions)
total_transitions += len(transitions)
total_assays += sum([collider.choose(len(transitions), i) for i in range(1, 6) if len(transitions) >= i])
progressm.update(1)
print "Total number of precursors is ", len(alltuples)
print "Total number of transitions is ", total_transitions
print "Total number of assays is ", total_assays
print "Obtained all peptides"
alltuples = list(cursor.fetchall() )
progressm = progress.ProgressMeter(total=len(alltuples), unit='peptides')
prepare = []
total_transitions = 0
total_assays = 0
for kk, pep in enumerate(alltuples):
q3_low, q3_high = [400, 1200]
q3_low, q3_high = [400, 1400]
q3charges = [1,2]
q3charges = [1]
#q3_low, q3_high = [0, 12000]
#
#new way to calculate the precursors
transitions = c_getnonuis.calculate_transitions_ch(
((-2, pep[0], -1),), q3charges, q3_low, q3_high)
#print pep, len(pep[0]), len(transitions)
total_transitions += len(transitions)
total_assays += sum([ collider.choose(len(transitions), i) for i in range(1,6)
if len(transitions) >= i ] )
progressm.update(1)
print "Total number of precursors is ", len(alltuples)
print "Total number of transitions is ", total_transitions
print "Total number of assays is ", total_assays
precursor, precursor.q1 - par.q1_window, precursor.q1 + par.q1_window,
transitions, par, rtree)
non_uis_list = collider.get_nonuis_list(collisions_per_peptide, MAX_UIS)
##
## Lets count the number of peptides that interfere
if count_avg_transitions:
tr_arr = [0 for i in range(nr_transitions)]
for v in collisions_per_peptide.values():
for vv in v:
tr_arr[vv] += 1
allintertr.extend( tr_arr )
##
#
for order in range(1,min(MAX_UIS+1, nr_transitions+1)):
prepare.append( (len(non_uis_list[order]), collider.choose(nr_transitions,
order), precursor.parent_id , order, exp_key) )
progressm.update(1)
if count_avg_transitions:
print "\n"
print "found %s transitions" % len(allintertr)#
print "found max of %s interferences" % max(allintertr)
print "found average of %s interferences" % ( sum(allintertr) * 1.0 / len(allintertr) )
# if any problems with the packet/buffer length occur, try this:
## set global max_allowed_packet=1000000000;
## set global net_buffer_length=1000000;
# cursor.executemany('insert into %s' % restable + ' (non_useable_UIS, total_UIS, \
# parent_key, uisorder, exp_key) values (%s,%s,%s,%s,%s)' , prepare)
print "Analyzed %s peptides" % len(precursors_to_evaluate)
###############################################################
#strike 1: it has to be global UIS
computed_collisions = myprecursors.get_collisions_per_peptide_from_rangetree(
precursor, precursor.q1 - par.q1_window, precursor.q1 + par.q1_window,
transitions, par, rtree)
collisions_per_peptide = computed_collisions
# see SRMCollider::Combinatorics::get_non_uis
non_useable_combinations = c_getnonuis.get_non_uis( collisions_per_peptide, myorder)
srm_ids = [t[1] for t in transitions]
tuples_strike1 = 0
if not nr_transitions < myorder:
tuples_strike1 = collider.choose(nr_transitions, myorder ) - len(non_useable_combinations)
###############################################################
#strike 2: it has to be locally clean
if not skip_strike2:
ssrcalc_low = ssrcalc - par.ssrcalc_window + 0.001
ssrcalc_high = ssrcalc + par.ssrcalc_window - 0.001
precursor_ids = tuple(c_rangetree.query_tree( q1_low, ssrcalc_low,
q1_high, ssrcalc_high ) )
precursors = tuple([parentid_lookup[myid[0]] for myid in precursor_ids
#dont select myself
if parentid_lookup[myid[0]][2] != pep['transition_group']])
# collisions_per_peptide: dictionary, for each key the set of interfering transitions is stored
collisions_per_peptide = c_getnonuis.calculate_collisions_per_peptide_other_ion_series(
transitions, precursors, par, q3_low, q3_high, par.q3_window, par.ppm, forceChargeCheck)
for precursor in precursors_to_evaluate:
transitions = precursor.calculate_transitions_from_param(par)
#correct rounding errors, s.t. we get the same results as before!
ssrcalc_low = precursor.ssrcalc - par.ssrcalc_window + 0.001
ssrcalc_high = precursor.ssrcalc + par.ssrcalc_window - 0.001
try:
result = c_integrated.wrap_all_bitwise(transitions,
precursor.q1 - par.q1_window, ssrcalc_low, precursor.q1 + par.q1_window, ssrcalc_high,
precursor.transition_group, min(par.max_uis,len(transitions)), par.q3_window, par.ppm,
par.isotopes_up_to, isotope_correction, par, r_tree)
except ValueError:
print "Too many transitions for", precursor.modification
continue
for order in range(1,min(par.max_uis+1, len(transitions)+1)):
prepare.append( (result[order-1], collider.choose(len(transitions),
order), precursor.parent_id , order, exp_key) )
#//break;
progressm.update(1)
for order in range(1,6):
sum_all = sum([p[0]*1.0/p[1] for p in prepare if p[3] == order])
nr_peptides = len([p for p in prepare if p[3] == order])
if not par.quiet and not nr_peptides ==0: print "Order %s, Average non useable UIS %s" % (order, sum_all *1.0/ nr_peptides)
#cursor.execute("insert into hroest.result_completegraph_aggr (sum_nonUIS, nr_peptides, uisorder, experiment) VALUES (%s,%s,%s,'%s')" % (sum_all, nr_peptides, order, exp_key))
"""
create table hroest.result_completegraph (
exp_key int(11),
parent_key int(11),
non_useable_UIS int(11),