def regression(points):
"""
do linear regression of all points
"""
sum_x = 0.00
sum_y = 0.00
sum_xx = 0.00
sum_xy = 0.00
sum_yy = 0.00
number_of_points = len(points)
for x, y in points:
pka_print("%6.2lf %6.2lf" % (x, y))
sum_x += x
sum_y += y
sum_xx += x * x
sum_xy += x * y
sum_yy += y * y
slope = (number_of_points * sum_xy -
sum_x * sum_y) / (number_of_points * sum_xx - sum_x * sum_x)
intercept = (sum_y * sum_xx -
sum_x * sum_xy) / (number_of_points * sum_xx - sum_x * sum_x)
correlation = (number_of_points * sum_xy - sum_x * sum_y) / math.sqrt(
(number_of_points * sum_xx - sum_x * sum_x) *
(number_of_points * sum_yy - sum_y * sum_y))
pka_print("%6.2lf + %8.4lf*x, r=%6.3lf" % (intercept, slope, correlation))
def compareWithExperiment(protein,
file=None,
list=None,
set=None,
label="ALL",
**argv):
"""
Compares the calculated pKa values with experiment if it is found in 'experiment'
"""
str = "# %s\n" % (protein.name)
file.write(str)
experimental = data.getExperiment(set=set, name=protein.name)
for chain in protein.chains:
for residue in chain.residues:
key = residue.label
if label == key or label == "ALL":
if key in experimental:
pka_exp = experimental[key]
dpka_exp = pka_exp - residue.pKa_mod
dpka_clc = residue.pKa_pro - residue.pKa_mod
diff = residue.pKa_pro - pka_exp
pka_print("compare:%s%6.2lf%6.2lf %6.2lf" %
(key, residue.pKa_pro, pka_exp, diff))
# adding to 'compare' list
list.append(diff)
# writing to 'comp.dat' file
str = "%8.2lf%6.2lf\n" % (dpka_exp, dpka_clc)
file.write(str)
def makeErrorPlot(points, filename=None):
"""
do patented error graph data
"""
# open file if needed
if filename != None:
file = open(filename, 'w')
# setup error-list
error_list = []
for i in range(0, 301):
error_list.append(0.00)
# loop over points and error bins
for point in points:
abs_diff = abs(point)
i = 0
while abs_diff > float(i)/100.0 and i < 301:
error_list[i] += 1.0
i += 1
number_of_points = float(len(points))
# print/write out error fraction
for i in range(len(error_list)):
error = float(i)/100.0
fraction = error_list[i]/number_of_points
str = "%6.2lf%6.3lf" % (error, fraction)
if filename == None:
pka_print(str)
else:
str += "\n"; file.write(str)
if filename != None:
file = open(filename, 'w')
def calculateRMSD(points):
"""
do rmsd of all points
"""
number_of_points = len(points)
sqr_sum = 0.00
for point in points:
sqr_sum += pow(point, 2)
rmsd = pow(sqr_sum / number_of_points, 0.5)
pka_print("rmsd = %6.4lf (%d)" % (rmsd, number_of_points))
def calculateMUE(points):
"""
do mean unsigned error of all points
"""
number_of_points = len(points)
abs_sum = 0.00
for point in points:
abs_sum += abs(point)
mue = abs_sum / number_of_points
pka_print("mue = %6.4lf" % (mue))
def calculateRMSD(points):
"""
do rmsd of all points
"""
number_of_points = len(points)
sqr_sum = 0.00
for point in points:
sqr_sum += pow(point, 2)
rmsd = pow(sqr_sum/number_of_points, 0.5)
pka_print("rmsd = %6.4lf (%d)" % (rmsd, number_of_points))
def calculateMUE(points):
"""
do mean unsigned error of all points
"""
number_of_points = len(points)
abs_sum = 0.00
for point in points:
abs_sum += abs(point)
mue = abs_sum/number_of_points
pka_print("mue = %6.4lf" % (mue))
def makeErrorPlot(points):
"""
do patented error graph data
"""
error_list = []
for i in range(0, 41):
error_list.append(0.00)
for point in points:
abs_diff = abs(point)
i = 0
while abs_diff > float(i)/10.0 and i < 41:
error_list[i] += 1.0
i += 1
number_of_points = float(len(points))
for i in range(0, 41):
error = float(i)/10.0
fraction = error_list[i]/number_of_points
pka_print("%6.2lf%6.2lf" % (error, fraction))
def printDesolvationDistances(protein):
"""
printing out wanted distances
"""
label = "GLU 66 A"
sum = 0.00
enumerator = 130.0
center_residue = protein.getResidue(label=label)
for chain in protein.chains:
for residue in chain.residues:
distance = 999.
closest = None
for atom in residue.atoms:
dX = atom.x - center_residue.x
dY = atom.y - center_residue.y
dZ = atom.z - center_residue.z
test_distance = math.sqrt(dX * dX + dY * dY + dZ * dZ)
if residue.label in [
"THR 62 A", "VAL 23 A", "LEU 14 A", "VAL 99 A",
"ILE 92 A", "LEU 36 A"
]:
if atom.element == "C" and atom.name not in ["C", "CA"]:
contribution = enumerator / pow(test_distance, 4)
sum += contribution
pka_print("%s %-3s %6.2lf %6.2lf" %
(residue.label, atom.name, test_distance,
contribution))
if test_distance < distance and atom.element == "C":
distance = test_distance
closest = atom
if distance < 8.0 and False:
pka_print("%s %6.2lf %s" %
(residue.label, distance, closest.name))
pka_print("sum:%8.4lf" % (sum))
def printDesolvationDistances(protein):
"""
printing out wanted distances
"""
label = "GLU 66 A"
sum = 0.00; enumerator = 130.0
center_residue = protein.getResidue(label=label)
for chain in protein.chains:
for residue in chain.residues:
distance = 999.; closest = None
for atom in residue.atoms:
dX = atom.x - center_residue.x
dY = atom.y - center_residue.y
dZ = atom.z - center_residue.z
test_distance = math.sqrt(dX*dX + dY*dY + dZ*dZ)
if residue.label in ["THR 62 A", "VAL 23 A", "LEU 14 A", "VAL 99 A", "ILE 92 A", "LEU 36 A"]:
if atom.element == "C" and atom.name not in ["C", "CA"]:
contribution = enumerator/pow(test_distance, 4)
sum += contribution
pka_print("%s %-3s %6.2lf %6.2lf" % (residue.label, atom.name, test_distance, contribution))
if test_distance < distance and atom.element == "C":
distance = test_distance
closest = atom
if distance < 8.0 and False:
pka_print("%s %6.2lf %s" % (residue.label, distance, closest.name))
pka_print("sum:%8.4lf" % (sum))
def compareWithExperiment(protein, file=None, list=None, set=None, label="ALL", **argv):
"""
Compares the calculated pKa values with experiment if it is found in 'experiment'
"""
str = "# %s\n" % (protein.name)
file.write(str)
experimental = data.getExperiment(set=set, name=protein.name)
for chain in protein.chains:
for residue in chain.residues:
key = residue.label
if label == key or label == "ALL":
if key in experimental:
pka_exp = experimental[key]
dpka_exp = pka_exp - residue.pKa_mod
dpka_clc = residue.pKa_pro - residue.pKa_mod
diff = residue.pKa_pro - pka_exp
pka_print( "compare:%s%6.2lf%6.2lf %6.2lf" % (key, residue.pKa_pro, pka_exp, diff) )
# adding to 'compare' list
list.append(diff)
# writing to 'comp.dat' file
str = "%8.2lf%6.2lf\n" % (dpka_exp, dpka_clc)
file.write(str)
def regression(points):
"""
do linear regression of all points
"""
sum_x = 0.00
sum_y = 0.00
sum_xx = 0.00
sum_xy = 0.00
sum_yy = 0.00
number_of_points = len(points)
for x, y in points:
pka_print("%6.2lf %6.2lf" % (x, y))
sum_x += x
sum_y += y
sum_xx += x*x
sum_xy += x*y
sum_yy += y*y
slope = (number_of_points*sum_xy - sum_x*sum_y) / (number_of_points*sum_xx - sum_x*sum_x)
intercept = (sum_y*sum_xx - sum_x*sum_xy) / (number_of_points*sum_xx - sum_x*sum_x)
correlation = (number_of_points*sum_xy - sum_x*sum_y) / math.sqrt((number_of_points*sum_xx - sum_x*sum_x)*(number_of_points*sum_yy - sum_y*sum_y))
pka_print("%6.2lf + %8.4lf*x, r=%6.3lf" % (intercept, slope, correlation))
def makeErrorPlot(points, filename=None):
"""
do patented error graph data
"""
# open file if needed
if filename != None:
file = open(filename, 'w')
# setup error-list
error_list = []
for i in range(0, 301):
error_list.append(0.00)
# loop over points and error bins
for point in points:
abs_diff = abs(point)
i = 0
while abs_diff > float(i) / 100.0 and i < 301:
error_list[i] += 1.0
i += 1
number_of_points = float(len(points))
# print/write out error fraction
for i in range(len(error_list)):
error = float(i) / 100.0
fraction = error_list[i] / number_of_points
str = "%6.2lf%6.3lf" % (error, fraction)
if filename == None:
pka_print(str)
else:
str += "\n"
file.write(str)
if filename != None:
file = open(filename, 'w')
def getExperiment(fullname):
"""
return experimental values
"""
name = fullname[:4]
if name == "1pga":
# B1_ProteinG
experiment = {
"GLU 15 A": 4.40,
"GLU 19 A": 3.70,
"ASP 22 A": 2.90,
"GLU 27 A": 4.50,
"ASP 36 A": 3.80,
"ASP 40 A": 4.00,
"GLU 42 A": 4.40,
"ASP 46 A": 3.60,
"ASP 47 A": 3.40,
"GLU 56 A": 4.00,
}
elif name == "1igd":
# B2_ProteinG
experiment = {
"GLU 20 A": 4.30,
"ASP 27 A": 2.90,
"GLU 29 A": 4.20,
"GLU 32 A": 4.60,
"ASP 41 A": 3.90,
"ASP 45 A": 4.40,
"ASP 51 A": 3.60,
"ASP 52 A": 3.40,
"GLU 61 A": 4.20,
}
elif name == "1a2p":
# Barnase
experiment = {
"ASP 8 A": 2.90,
"ASP 12 A": 3.80,
"ASP 22 A": 3.30,
"GLU 29 A": 3.80,
"ASP 44 A": 3.40,
# "ASP 54 A": 2.20, # <= 2.2
"GLU 60 A": 3.00,
# "GLU 73 A": 2.10, # <= 2.1
# "ASP 75 A": 3.10, # difficult barnase pi-stacking
"ASP 86 A": 4.20,
# "ASP 93 A": 2.00, # < 2.0
# "ASP 101 A": 2.00, # <= 2
}
elif name == "1rnz":
# Bovine_RNase
experiment = {
"GLU 2 A": 2.70,
"GLU 9 A": 4.00,
# "ASP 14 A": 1.80, # < 2.0
"ASP 38 A": 2.80,
"GLU 49 A": 4.50,
"ASP 53 A": 3.80,
"ASP 83 A": 3.40,
"GLU 86 A": 4.05,
"GLU 111 A": 3.50,
"ASP 121 A": 3.05,
}
elif name == "4icb":
# Calbindin
experiment = {
"GLU 4 A": 3.80,
"GLU 5 A": 3.40,
"GLU 11 A": 4.70,
"GLU 17 A": 3.60,
"GLU 26 A": 4.10,
"ASP 47 A": 3.00,
"GLU 48 A": 4.60,
"GLU 64 A": 3.80,
}
elif name == "1kxi":
# Cardiotoxin
experiment = {
"GLU 17 A": 4.00,
"ASP 42 A": 3.20,
# "ASP 59 A": 2.30, # < 2.3
"GLU 17 B": 4.00,
"ASP 42 B": 3.20,
# "ASP 59 B": 2.30, # < 2.3
}
elif name == "1kxi_A":
# Cardiotoxin
experiment = {"GLU 17 A": 4.00,
"ASP 42 A": 3.20,
"ASP 59 A": 2.30}
elif name == "1kxi_B":
# Cardiotoxin
experiment = {"GLU 17 B": 4.00,
"ASP 42 B": 3.20,
"ASP 59 B": 2.30}
elif name == "1cdc":
# CD2d1
experiment = {
"ASP 25 A": 3.50,
"ASP 26 A": 3.60,
"ASP 28 A": 3.60,
"GLU 29 A": 4.40,
"GLU 33 A": 4.20,
"GLU 41 A": 6.70,
"GLU 56 A": 3.90,
"ASP 62 A": 4.20,
"ASP 71 A": 3.20,
"ASP 72 A": 4.10,
"ASP 94 A": 3.90,
"GLU 99 A": 4.20,
"ASP 25 B": 3.50,
"ASP 26 B": 3.60,
"ASP 28 B": 3.60,
"GLU 29 B": 4.40,
"GLU 33 B": 4.20,
"GLU 41 B": 6.70,
"GLU 56 B": 3.90,
"ASP 62 B": 4.20,
"ASP 71 B": 3.20,
"ASP 72 B": 4.10,
"ASP 94 B": 3.90,
"GLU 99 B": 4.20,
}
elif name == "1beo":
# Cryptogein
experiment = {
"ASP 21 A": 2.50,
"ASP 30 A": 2.50,
"ASP 72 A": 2.60,
}
elif name == "1hpx":
# HIV_protease, note, homodimer, B is assumed to be A
experiment = {
#"ASP 25 B": 6.20, # > 6.2
"ASP 29 B": 3.20,
"ASP 30 B": 3.90,
"ASP 60 B": 3.00,
#"ASP 25 A": 2.50, # < 2.5
"ASP 29 A": 3.70,
"ASP 30 A": 3.80,
"ASP 60 A": 3.00,
}
elif name == "1lys":
# Hen lysozyme
experiment = {
"GLU 7 A": 2.90,
"ASP 18 A": 2.70,
"GLU 35 A": 6.20,
# "ASP 48 A": 2.50, # < 2.5
"ASP 52 A": 3.70,
# "ASP 66 A": 2.00, # < 2.5
"ASP 87 A": 2.10,
"ASP 101 A": 4.10,
"ASP 119 A": 3.20,
}
elif name == "135l":
# Turkey lysozyme
experiment = {
"GLU 7 A": 2.70,
"ASP 18 A": 2.70,
"GLU 35 A": 6.10,
# "ASP 48 A": 2.50, # < 2.5
"ASP 52 A": 3.80,
# "ASP 66 A": 2.00, # < 2.5
"ASP 87 A": 2.10,
"ASP 119 A": 3.40,
}
elif name == "2rn2":
# Ribonuclease H1
experiment = {
"GLU 6 A": 4.50,
"ASP 10 A": 6.10,
"GLU 32 A": 3.60,
"GLU 48 A": 4.40,
"GLU 57 A": 3.20,
"GLU 61 A": 3.90,
"GLU 64 A": 4.40,
"ASP 70 A": 2.60,
"ASP 94 A": 3.20,
# "ASP 102 A": 2.00, # < 2
"ASP 108 A": 3.20,
"GLU 119 A": 4.10,
"GLU 129 A": 3.60,
"GLU 131 A": 4.30,
"ASP 134 A": 4.10,
"GLU 135 A": 4.30,
"GLU 147 A": 4.20,
# "ASP 148 A": 2.00, # < 2
"GLU 154 A": 4.40,
}
elif name == "2ovo":
# Ovomucoid3Dom
experiment = {
"ASP 7 A": 2.50,
"GLU 10 A": 4.10,
"GLU 19 A": 3.20,
"ASP 27 A": 2.20,
"GLU 43 A": 4.80,
}
elif name == "1xnb":
# Xylanase
experiment = {
"ASP 4 A": 3.00,
"ASP 11 A": 2.50,
"GLU 78 A": 4.60,
# "ASP 83 A": 2.00, # < 2
# "ASP 101 A": 2.00, # < 2
"ASP 106 A": 2.70,
"ASP 119 A": 3.20,
"ASP 121 A": 3.60,
"GLU 172 A": 6.70,
}
elif name == "1de3":
# Sarcin
experiment = {
"ASP 9 A": 3.90,
"GLU 19 A": 4.60,
"GLU 31 A": 4.60,
# "ASP 41 A": 3.00, # < 3
"ASP 57 A": 4.30,
"ASP 59 A": 4.10,
"ASP 75 A": 3.90,
# "ASP 77 A": 3.00, # < 3
"ASP 85 A": 3.80,
# "ASP 91 A": 3.00, # < 3
"GLU 96 A": 5.10,
# "ASP 102 A": 3.00, # < 3
# "ASP 105 A": 3.00, # < 3
"ASP 109 A": 3.70,
"GLU 115 A": 4.90,
"GLU 140 A": 4.30,
"GLU 144 A": 4.30,
}
elif name == "2bus":
# BullSeminalInhibitor
experiment = {
"ASP 6 A": 4.00,
"GLU 9 A": 4.30,
"ASP 12 A": 3.60,
"GLU 20 A": 4.10,
}
elif name == "1egf":
# Epidermal Growthfactor
experiment = {
"ASP 11 A": 3.90,
"GLU 24 A": 4.10,
"ASP 27 A": 4.00,
"ASP 40 A": 3.60,
"ASP 46 A": 3.80,
"GLU 51 A": 4.00,
}
elif name == "1mhi":
# Insulin
experiment = {
"GLU 4 A": 2.60,
# "GLU 17 A": 3.70, # > 3.7
"ASP 9 B": 2.60,
"GLU 13 B": 2.20,
"GLU 21 B": 3.70,
}
else:
pka_print("could not find experimental data for \"%s\"" % (name))
sys.exit(9)
return experiment