def do_dir(path,detailed=None):
#
# Load experimental pKa values
#
expdata='/u1/jnielsen/work/pka/exppka/lysozyme.PKA.DAT'
experimental=pKa.pKaIO()
experimental.readpka(expdata)
A=pKa.pKanalyse()
#
# Find all directories with pKa calculations and loop over them
#
list=ostools.find('*.PKA.DAT',path)
topdir=os.getcwd()
print
print '================================================'
print 'Evaluating ',len(list),' calculations in: %s' %os.path.split(path)[1]
#
# Reset counters
#
correct=0
wrong=0
rmsds=[]
E35sum=[]
D52sum=[]
for file in list:
dir,pkafile=os.path.split(file)
pos=string.find(dir,path)
os.chdir(dir)
if pos!=1:
dir=dir[pos+len(path):]
calculated=pKa.pKaIO()
calculated.readpka(pkafile)
rmsd=A.rmsd_between_sets(experimental.pka,calculated.pka)
if detailed:
print 'Dir: %6s rmsd: %5.2f ' %(dir,rmsd)
#
# Get stats on D52 and E35
#
residues=['0035GLU','0052ASP']
print ' Residue Exp. Calc. diff'
for residue in residues:
print ' %15s %5.2f %5.2f %5.2f' %(residue,experimental.pka[residue]['pKa'], calculated.pka[residue]['pKa'],experimental.pka[residue]['pKa']-calculated.pka[residue]['pKa'])
if calculated.pka['0035GLU']['pKa']-calculated.pka['0052ASP']['pKa']>=1.5 and calculated.pka['0035GLU']['pKa']>5.0:
print 'GLU35!'
elif -calculated.pka['0035GLU']['pKa']+calculated.pka['0052ASP']['pKa']>=1.5 and calculated.pka['0052ASP']['pKa']>5.0:
print 'ASP52!'
else:
print 'None'
print
#
# Did we get the h-donor right?
#
if calculated.pka.has_key('0035GLU'):
E35=calculated.pka['0035GLU']['pKa']
E35sum.append(E35)
if calculated.pka.has_key('0052ASP'):
D52=calculated.pka['0052ASP']['pKa']
D52sum.append(D52)
if calculated.pka.has_key('0035GLU') and calculated.pka.has_key('0052ASP'):
if E35>5.0 and E35-D52>=1.5:
correct=correct+1
if D52>5.0 and D52-E35>=1.5:
wrong=wrong+1
rmsds.append(rmsd)
#
# --------------------------------------------------------
#
# Print the summary
#
print 'Number of dirs',len(rmsds)
if len(rmsds)>0:
sum=0.0
for rmsd in rmsds:
sum=sum+rmsd
Esum=0.0
for p in E35sum:
Esum=Esum+p
Dsum=0.0
for p in D52sum:
Dsum=Dsum+p
D52mean=Dsum/float(len(D52sum))
E35mean=Esum/float(len(E35sum))
avgdevE=0.0
avgdevD=0.0
for p in E35sum:
avgdevE=avgdevE+abs(p-E35mean)
for p in D52sum:
avgdevD=avgdevD+abs(p-D52mean)
avgdevE=avgdevE/float(len(E35sum))
avgdevD=avgdevD/float(len(D52sum))
print 'Average rmsd: %5.2f' %(sum/float(len(rmsds)))
print 'Average pKa E35: %5.2f' %(Esum/float(len(E35sum)))
print 'Average pKa D52: %5.2f' %(Dsum/float(len(D52sum)))
print 'Avg. deviation from mean (E35): %7.3f' %avgdevE
print 'Avg. deviation from mean (D52): %7.3f' %avgdevD
print 'Correct:',correct
print 'Wrong:',wrong
#
# Average the titration curves
#
list=ostools.find('*.TITCURV.DAT',path)
avg_charge_pKa(list)
os.chdir(topdir)
return
def avg_charge_pKa(files):
#
# Calculate the average titration curve and the average pKa for the dirs
#
import sys, os, string
#dirs=sys.argv[1:]
top=os.getcwd()
pkadict={}
titdict={}
ok=0
wrong=0
for file in files:
if os.path.isfile(file):
#print 'Processing file:',file
import pKa
if os.path.isfile(file):
X=pKa.pKaIO()
tit=X.readtitcurv(file)
titdict[file]={}
if tit.has_key('0035GLU') and tit.has_key('0052ASP'):
titdict[file]={'35':tit['0035GLU'],'52':tit['0052ASP']}
else:
print 'pKa calculations not finished for: %s' %dir
#
# Calculate averages
#
sumtit={}
#
# Zero dictionaries
#
for outkey in titdict.keys():
for key in titdict[outkey].keys():
sumtit[key]={}
#
# Sum all
#
pkacount=0
for dir in titdict.keys():
for key in titdict[dir].keys():
for ph in titdict[dir][key].keys():
if sumtit[key].has_key(ph):
sumtit[key][ph]=sumtit[key][ph]+titdict[dir][key][ph]
else:
sumtit[key][ph]=titdict[dir][key][ph]
#
# Get average
#
for key in sumtit.keys():
for ph in sumtit[key].keys():
sumtit[key][ph]=float(sumtit[key][ph])/float(len(titdict.keys()))
#
# Print the results
#
residues=sumtit.keys()
residues.sort()
pka=0.0
print 'pKa value from average titration curve'
for key in residues:
phvals=sumtit[key].keys()
phvals.sort()
for ph in phvals:
if sumtit[key][ph]<=-0.5:
pka=ph
break
print "Residue: %s, pKa value: %5.2f" %(key,pka)
return
def main():
#
# Get the PDB file
#
import sys,os
pdbfile=sys.argv[1]
suffix=sys.argv[2]
if os.environ['USER']=='nielsen':
filename=os.path.join('/enzyme/nielsen/work/pKa_design/accuracy','accuracy_'+os.path.split(pdbfile)[1])
elif os.environ['USER']=='btconnolly':
filename=os.path.join('/enzyme/btconnolly/pKa_design',suffix+'accuracy_'+os.path.split(pdbfile)[1])
#filename=os.path.join(os.getcwd(),'accuracy_'+os.path.split(pdbfile)[1])
print 'Setting dictionary filename to',filename
#
# Make sure that we delete all info from previous runs
#
wtpKafile=pdbfile+'_wt_pKavals'
if os.path.isfile(wtpKafile):
os.unlink(wtpKafile)
#
# Do we have a completed pKa calc for it?
#
import pKa
X=pKa.pKaIO(pdbfile)
X.usenewnames()
if not X.assess_status():
import os
print 'You have to run a pKa calculation first'
raise Exception()
#
# OK, got the pKa calc. Read results
#
wt_pKas=X.readpka()
groups=wt_pKas.keys()
groups.sort()
#
# Design target: for all groups with pKa value between 2 and 10: +2 and -2
#
DB=dictIO(filename)
import os
if os.path.isfile(filename):
#
# Load old file
#
accuracy=DB.load()
else:
#
# No, no old restuls
#
accuracy={}
#
# Add the PDB file
#
fd=open(pdbfile)
lines=fd.readlines()
fd.close()
accuracy['pdbfile']=lines
#
# Add the full wt pKa values
#
accuracy['wt_full']=wt_pKas
DB.save(accuracy)
# ---------------------------------------
#
# Delete old files
#
import Design_pKa
Design_pKa.delete_old_files(pdbfile,'N','N')
#
# Start looping
#
groups.sort()
for group in groups:
#if not group==':0129:LEU:CTERM':
# continue
if not accuracy.has_key(group):
accuracy[group]={}
#
# Is somebody working on this group?
#
accuracy=DB.update(accuracy)
if accuracy[group].has_key('locked'):
if accuracy[group]['locked']==1:
continue
#
# Lock group
#
accuracy[group]['locked']=1
#
# Save the dictionary
#
accuracy=DB.update(accuracy)
#
# OK, now we can calculate in peace
#
pKaval=wt_pKas[group]['pKa']
if pKaval>2.0 and pKaval<10.0:
#
# First design pKa +2.0
#
design='+2.0'
if not accuracy[group].has_key(design):
accuracy[group][design]={}
print 'Designing and evalulating: %s' %(group+design)
dpKas=get_solutions(pdbfile,group,design,accuracy[group][design])
accuracy[group][design]=dpKas.copy()
#
# Then do pKa -2.0
#
design='m2.0'
if not accuracy[group].has_key(design):
accuracy[group][design]={}
print 'Designing and evalulating: %s' %(group+design)
dpKas=get_solutions(pdbfile,group,design,accuracy[group][design])
accuracy[group][design]=dpKas.copy()
else:
accuracy[group]['pKa out of range']=1
#
# Unlock group and merge results
#
accuracy[group]['locked']=None
accuracy=DB.update(accuracy)
#
# All done
#
return
def main():
#
# Get the PDB file
#
import sys, os
pdbfile = sys.argv[1]
suffix = sys.argv[2]
if os.environ['USER'] == 'nielsen':
filename = os.path.join('/enzyme/nielsen/work/pKa_design/accuracy',
'accuracy_' + os.path.split(pdbfile)[1])
elif os.environ['USER'] == 'btconnolly':
filename = os.path.join(
'/enzyme/btconnolly/pKa_design',
suffix + 'accuracy_' + os.path.split(pdbfile)[1])
#filename=os.path.join(os.getcwd(),'accuracy_'+os.path.split(pdbfile)[1])
print 'Setting dictionary filename to', filename
#
# Make sure that we delete all info from previous runs
#
wtpKafile = pdbfile + '_wt_pKavals'
if os.path.isfile(wtpKafile):
os.unlink(wtpKafile)
#
# Do we have a completed pKa calc for it?
#
import pKa
X = pKa.pKaIO(pdbfile)
X.usenewnames()
if not X.assess_status():
import os
print 'You have to run a pKa calculation first'
raise Exception()
#
# OK, got the pKa calc. Read results
#
wt_pKas = X.readpka()
groups = wt_pKas.keys()
groups.sort()
#
# Design target: for all groups with pKa value between 2 and 10: +2 and -2
#
DB = dictIO(filename)
import os
if os.path.isfile(filename):
#
# Load old file
#
accuracy = DB.load()
else:
#
# No, no old restuls
#
accuracy = {}
#
# Add the PDB file
#
fd = open(pdbfile)
lines = fd.readlines()
fd.close()
accuracy['pdbfile'] = lines
#
# Add the full wt pKa values
#
accuracy['wt_full'] = wt_pKas
DB.save(accuracy)
# ---------------------------------------
#
# Delete old files
#
import Design_pKa
Design_pKa.delete_old_files(pdbfile, 'N', 'N')
#
# Start looping
#
groups.sort()
for group in groups:
#if not group==':0129:LEU:CTERM':
# continue
if not accuracy.has_key(group):
accuracy[group] = {}
#
# Is somebody working on this group?
#
accuracy = DB.update(accuracy)
if accuracy[group].has_key('locked'):
if accuracy[group]['locked'] == 1:
continue
#
# Lock group
#
accuracy[group]['locked'] = 1
#
# Save the dictionary
#
accuracy = DB.update(accuracy)
#
# OK, now we can calculate in peace
#
pKaval = wt_pKas[group]['pKa']
if pKaval > 2.0 and pKaval < 10.0:
#
# First design pKa +2.0
#
design = '+2.0'
if not accuracy[group].has_key(design):
accuracy[group][design] = {}
print 'Designing and evalulating: %s' % (group + design)
dpKas = get_solutions(pdbfile, group, design,
accuracy[group][design])
accuracy[group][design] = dpKas.copy()
#
# Then do pKa -2.0
#
design = 'm2.0'
if not accuracy[group].has_key(design):
accuracy[group][design] = {}
print 'Designing and evalulating: %s' % (group + design)
dpKas = get_solutions(pdbfile, group, design,
accuracy[group][design])
accuracy[group][design] = dpKas.copy()
else:
accuracy[group]['pKa out of range'] = 1
#
# Unlock group and merge results
#
accuracy[group]['locked'] = None
accuracy = DB.update(accuracy)
#
# All done
#
return
def avg_charge_pKa(files):
#
# Calculate the average titration curve and the average pKa for the dirs
#
import sys, os, string
#dirs=sys.argv[1:]
top = os.getcwd()
pkadict = {}
titdict = {}
ok = 0
wrong = 0
for file in files:
if os.path.isfile(file):
#print 'Processing file:',file
import pKa
if os.path.isfile(file):
X = pKa.pKaIO()
tit = X.readtitcurv(file)
titdict[file] = {}
if tit.has_key('0035GLU') and tit.has_key('0052ASP'):
titdict[file] = {
'35': tit['0035GLU'],
'52': tit['0052ASP']
}
else:
print 'pKa calculations not finished for: %s' % dir
#
# Calculate averages
#
sumtit = {}
#
# Zero dictionaries
#
for outkey in titdict.keys():
for key in titdict[outkey].keys():
sumtit[key] = {}
#
# Sum all
#
pkacount = 0
for dir in titdict.keys():
for key in titdict[dir].keys():
for ph in titdict[dir][key].keys():
if sumtit[key].has_key(ph):
sumtit[key][ph] = sumtit[key][ph] + titdict[dir][key][ph]
else:
sumtit[key][ph] = titdict[dir][key][ph]
#
# Get average
#
for key in sumtit.keys():
for ph in sumtit[key].keys():
sumtit[key][ph] = float(sumtit[key][ph]) / float(
len(titdict.keys()))
#
# Print the results
#
residues = sumtit.keys()
residues.sort()
pka = 0.0
print 'pKa value from average titration curve'
for key in residues:
phvals = sumtit[key].keys()
phvals.sort()
for ph in phvals:
if sumtit[key][ph] <= -0.5:
pka = ph
break
print "Residue: %s, pKa value: %5.2f" % (key, pka)
return
def do_dir(path, detailed=None):
#
# Load experimental pKa values
#
expdata = '/u1/jnielsen/work/pka/exppka/lysozyme.PKA.DAT'
experimental = pKa.pKaIO()
experimental.readpka(expdata)
A = pKa.pKanalyse()
#
# Find all directories with pKa calculations and loop over them
#
list = ostools.find('*.PKA.DAT', path)
topdir = os.getcwd()
print
print '================================================'
print 'Evaluating ', len(
list), ' calculations in: %s' % os.path.split(path)[1]
#
# Reset counters
#
correct = 0
wrong = 0
rmsds = []
E35sum = []
D52sum = []
for file in list:
dir, pkafile = os.path.split(file)
pos = string.find(dir, path)
os.chdir(dir)
if pos != 1:
dir = dir[pos + len(path):]
calculated = pKa.pKaIO()
calculated.readpka(pkafile)
rmsd = A.rmsd_between_sets(experimental.pka, calculated.pka)
if detailed:
print 'Dir: %6s rmsd: %5.2f ' % (dir, rmsd)
#
# Get stats on D52 and E35
#
residues = ['0035GLU', '0052ASP']
print ' Residue Exp. Calc. diff'
for residue in residues:
print ' %15s %5.2f %5.2f %5.2f' % (
residue, experimental.pka[residue]['pKa'],
calculated.pka[residue]['pKa'],
experimental.pka[residue]['pKa'] -
calculated.pka[residue]['pKa'])
if calculated.pka['0035GLU']['pKa'] - calculated.pka['0052ASP'][
'pKa'] >= 1.5 and calculated.pka['0035GLU']['pKa'] > 5.0:
print 'GLU35!'
elif -calculated.pka['0035GLU']['pKa'] + calculated.pka['0052ASP'][
'pKa'] >= 1.5 and calculated.pka['0052ASP']['pKa'] > 5.0:
print 'ASP52!'
else:
print 'None'
print
#
# Did we get the h-donor right?
#
if calculated.pka.has_key('0035GLU'):
E35 = calculated.pka['0035GLU']['pKa']
E35sum.append(E35)
if calculated.pka.has_key('0052ASP'):
D52 = calculated.pka['0052ASP']['pKa']
D52sum.append(D52)
if calculated.pka.has_key('0035GLU') and calculated.pka.has_key(
'0052ASP'):
if E35 > 5.0 and E35 - D52 >= 1.5:
correct = correct + 1
if D52 > 5.0 and D52 - E35 >= 1.5:
wrong = wrong + 1
rmsds.append(rmsd)
#
# --------------------------------------------------------
#
# Print the summary
#
print 'Number of dirs', len(rmsds)
if len(rmsds) > 0:
sum = 0.0
for rmsd in rmsds:
sum = sum + rmsd
Esum = 0.0
for p in E35sum:
Esum = Esum + p
Dsum = 0.0
for p in D52sum:
Dsum = Dsum + p
D52mean = Dsum / float(len(D52sum))
E35mean = Esum / float(len(E35sum))
avgdevE = 0.0
avgdevD = 0.0
for p in E35sum:
avgdevE = avgdevE + abs(p - E35mean)
for p in D52sum:
avgdevD = avgdevD + abs(p - D52mean)
avgdevE = avgdevE / float(len(E35sum))
avgdevD = avgdevD / float(len(D52sum))
print 'Average rmsd: %5.2f' % (sum / float(len(rmsds)))
print 'Average pKa E35: %5.2f' % (Esum / float(len(E35sum)))
print 'Average pKa D52: %5.2f' % (Dsum / float(len(D52sum)))
print 'Avg. deviation from mean (E35): %7.3f' % avgdevE
print 'Avg. deviation from mean (D52): %7.3f' % avgdevD
print 'Correct:', correct
print 'Wrong:', wrong
#
# Average the titration curves
#
list = ostools.find('*.TITCURV.DAT', path)
avg_charge_pKa(list)
os.chdir(topdir)
return