def statsSummary(pdbSet, data, geos,tag):
import matplotlib.pyplot as plt
plt.close('all')
plt.clf()
plt.cla()
pdbDataPath = help.rootPath + '/ProteinDataFiles/pdb_out/' + pdbSet + '/'
edDataPath = '/home/rachel/Documents/Bioinformatics/ProteinDataFiles/ccp4_data/'
loadPath = help.rootPath + '/BbkProject/PhDThesis/0.Papers/3.DefensibleGeometry/EvidencedSet/DataB/'
printPath = help.rootPath + '/BbkProject/PhDThesis/0.Papers/3.DefensibleGeometry/EvidencedSet/DataK/'
fileName = 'Data_DefensibleWithGeosALL_' + pdbSet + '.csv'
allAtoms = False
bFactorFactor = -1
if pdbSet == 'RESTRICTED':
allAtoms = True
bFactorFactor = 1.3
georep = psu.GeoReport([], pdbDataPath, edDataPath, printPath, ed=False, dssp=False, includePdbs=False, keepDisordered=False)
for geo in geos:
georep.addStatsSummary(data=data, desc=geo + ' ' + pdbSet, geoX=geo, geoY='aa', hue='ID')
georep.printToHtml('Stats Summary , set=' + pdbSet, 2, 'StatsSummary_' + pdbSet + tag)
def diffHistograms(pdbList, tag):
pdbDataPath = help.rootPath + '/ProteinDataFiles/pdb_out/'
edDataPath = '/home/rachel/Documents/Bioinformatics/ProteinDataFiles/ccp4_data/'
loadPath = help.rootPath + '/ProteinDataFiles/ccp4_out/'
printPath = help.rootPath + '/BbkProject/PhDThesis/0.Papers/3.DefensibleGeometry/EvidencedSet/DataN/'
allRealPdbs = []
for pdb in pdbList:
realFileName = pdb + '_DiffHistogram.csv'
realData = pd.read_csv(loadPath + realFileName)
allRealPdbs.append(realData)
#append them all
realCsv = pd.concat(allRealPdbs, axis=0, sort=False)
georep = psu.GeoReport([],
pdbDataPath,
edDataPath,
printPath,
ed=False,
dssp=False,
includePdbs=False,
keepDisordered=False)
georep.addHistogram(data=realCsv,
geoX='Percent',
title='Difference Histograms - %',
count=True,
hue='PdbCode')
georep.addHistogram(data=realCsv,
geoX='Diff',
title='Difference Histograms - diff',
count=True,
hue='PdbCode',
palette='LightSeaGreen')
georep.addScatter(data=realCsv,
geoX='Main',
geoY='Percent',
hue='Diff',
palette='jet',
categorical=False,
sort='RANDOM',
title='Differences')
georep.addScatter(data=realCsv,
geoX='Diff',
geoY='Percent',
hue='PdbCode',
palette='jet_r',
categorical=True,
sort='RANDOM',
title='Differences')
georep.printToHtml('Difference Histograms', 2, 'DiffHist_' + tag)
def scatterReports(pdbSet, data, trios, perAA=True, tag=''):
import matplotlib.pyplot as plt
plt.close('all')
plt.clf()
plt.cla()
pdbDataPath = help.rootPath + '/ProteinDataFiles/pdb_data/'
edDataPath = '/home/rachel/Documents/Bioinformatics/ProteinDataFiles/ccp4_data/'
printPath = help.rootPath + '/BbkProject/PhDThesis/0.Papers/3.DefensibleGeometry/EvidencedSet/DataI/'
#BestFileName = 'Data_DefensibleWithGeosALL_' + pdbSet + '.csv'
#dataBest = pd.read_csv(loadPath + BestFileName)
aas = data['aa'].values
aas = list(set(aas))
aas.sort()
#aas = ['ALA', 'CYS', 'ASP', 'GLU', 'PHE', 'GLY', 'HIS', 'ILE', 'LYS', 'LEU', 'MET', 'ASN', 'PRO', 'GLN', 'ARG','SER', 'THR', 'VAL', 'TRP', 'TYR']
#geosPairs = [['PHI','PSI','TAU'],['PSI','N:N+1','TAU'],['N:CA','CA:C','C:O'],['C:N+1','TAU','CA:C:N+1'],['CA:C:O','O:C:N+1','C-1:N:CA']]
georep = psu.GeoReport([],
pdbDataPath,
edDataPath,
printPath,
ed=False,
dssp=False,
includePdbs=False,
keepDisordered=False)
for trio in trios:
if perAA:
for aa in aas:
dataCut = data.query("aa == '" + aa + "'")
georep.addScatter(data=dataCut,
geoX=trio[0],
geoY=trio[1],
hue=trio[2],
title=aa + ':' + trio[0] + ':' + trio[1],
palette='jet',
sort='NON')
else:
georep.addScatter(data=data,
geoX=trio[0],
geoY=trio[1],
hue=trio[2],
title=trio[0] + ':' + trio[1],
palette='jet',
sort='NON')
georep.printToHtml('Scatters , set=' + pdbSet, 4,
'Defensible_Scatters_' + tag)
def makeSlicesHtmlFromValues(mainTitle, dirName, lineRuns,row,tag):
printPath = help.rootPath + '/BbkProject/PhDThesis/0.Papers/3.DefensibleGeometry/EvidencedSet/SlicesI/'
georep = psu.GeoReport([], "", "", printPath, ed=False, dssp=False)
#We are going to load the data that has been created by density flight and pasted into a text file
for lineRun in lineRuns:
title = lineRun[0]
palette = lineRun[1]
fileName = lineRun[2]
posName = lineRun[3]
inputVals = georep.loadSlice(dirName + fileName)
inputPoses = georep.loadSlice(dirName + posName)
georep.addSlice(inputVals, palette=palette, title=title, YellowDots=inputPoses,Contour=True)
georep.printToHtml(mainTitle,row,tag)
def __init__(self,plotA, plotB, title,report):
self.title = title
if title!='ghost':
self.plotA = plotA
self.plotB = plotB
else:#In this case we have only the main plot, so we create the dummy plot
from PsuGeometry import GeoReport as geor
self.plotB = plotA
ghostReport = geor.GeoReport(['ghost'],report.pdbDataPath,report.edDataPath,report.outDataPath,report.ed,report.dssp)
geoList = []
geoList.append(self.plotB.geoX)
if self.plotB.geoY != '':
geoList.append(self.plotB.geoY)
ghostdata = ghostReport.getGeoemtryCsv(geoList, ['pdbCode'])
self.plotA = GeoPlot(ghostdata, self.plotB.geoX, geoY=self.plotB.geoY, title='ghost', hue='pdbCode', palette='Greys',plot=self.plotB.plot,operation=self.plotB.operation,report=report)
def makeCsv(pdbSet, pdbListIn, geos, badAtoms, disordered):
print('Getting CSV for', pdbSet)
pdbDataPath = filesPDBRoot
if pdbSet == 'ADJUSTEDDEN':
pdbDataPath = filesDenRoot
if pdbSet == 'ADJUSTEDLAP':
pdbDataPath = filesLapRoot
from PsuGeometry import GeoPdb as geopdb
pdbmanager = geopdb.GeoPdbs(pdbDataPath, edDataPath, False, False,
disordered, badAtoms)
pdbmanager.clear()
pdbmanager = geopdb.GeoPdbs(pdbDataPath, edDataPath, False, False,
disordered, badAtoms)
pdbList = [
] #this is so we don't default to getting a pdb file from somewhere we don;t want
for pdb in pdbListIn:
import os.path
filePdb = pdbDataPath + 'pdb' + pdb + '.ent'
#print('- Adding to csv',filePdb)
if os.path.isfile((filePdb).lower()):
pdbList.append(pdb.lower())
else:
print('No file:', pdbDataPath, pdb)
pdbList.sort()
hueList = [
'aa', 'rid', 'bfactor', 'pdbCode', 'bfactorRatio', 'disordered',
'occupancy'
]
georep = psu.GeoReport(pdbList,
pdbDataPath,
edDataPath,
printPath,
ed=False,
dssp=False,
includePdbs=False,
keepDisordered=disordered)
print('geoList', geos)
dataBest = georep.getGeoemtryCsv(geos, hueList, -1, allAtoms=True)
try:
dataBest['rid'] = dataBest['rid'].astype(str)
dataBest['ID'] = dataBest['pdbCode'] + dataBest['chain'] + dataBest[
'rid'] + dataBest['aa']
except:
print('empty csv')
return dataBest
def clusterEdTauMaker(pdbCode, rid, chain, aa):
from PsuGeometry import GeoReport as psu
from PsuGeometry import GeoPdb as geopdb
pdbDataPath = '/home/rachel/Documents/Bioinformatics/ProteinDataFiles/pdb_data/'
edDataPath = '/home/rachel/Documents/Bioinformatics/ProteinDataFiles/ccp4_data/'
printPath = 'F:/Code/BbkProject/PhDThesis/0.Papers/1.TauCorrelations/Data/BestSupportedCSVs/Reports/'
georep = psu.GeoReport([pdbCode],
pdbDataPath,
edDataPath,
printPath,
ed=False,
dssp=False)
pdbmanager = geopdb.GeoPdbs(pdbDataPath, edDataPath, ed=False, dssp=False)
apdb = pdbmanager.getPdb(pdbCode, True)
pdbcsv = apdb.getDataFrame()
queryC = 'rid==' + str(
rid) + ' and chain=="' + chain + '"' + ' and atom=="CA"'
queryL = 'rid==' + str(
rid) + ' and chain=="' + chain + '"' + ' and atom=="N"'
queryP = 'rid==' + str(
rid) + ' and chain=="' + chain + '"' + ' and atom=="C"'
dataC = pdbcsv.query(queryC)
dataL = pdbcsv.query(queryL)
dataP = pdbcsv.query(queryP)
if len(dataC) > 0 and len(dataL) > 0 and len(dataP) > 0:
cx = round(dataC['x'].values[0], 3)
cy = round(dataC['y'].values[0], 3)
cz = round(dataC['z'].values[0], 3)
lx = round(dataL['x'].values[0], 3)
ly = round(dataL['y'].values[0], 3)
lz = round(dataL['z'].values[0], 3)
px = round(dataP['x'].values[0], 3)
py = round(dataP['y'].values[0], 3)
pz = round(dataP['z'].values[0], 3)
row = pdbCode + "," + chain + str(rid) + "," + str(cx) + "," + str(
cy) + "," + str(cz)
row += "," + str(lx) + "," + str(ly) + "," + str(lz)
row += "," + str(px) + "," + str(py) + "," + str(pz)
return row
def evidenceReports(pdbSet, fourSetNames, dataA, dataB, dataC, dataD ,trios, title,perAA=True, tag=''):
pdbDataPath = help.rootPath + '/ProteinDataFiles/pdb_data/'
edDataPath = '/home/rachel/Documents/Bioinformatics/ProteinDataFiles/ccp4_data/'
printPath = help.rootPath + '/BbkProject/PhDThesis/0.Papers/3.DefensibleGeometry/EvidencedSet/DataI/'
aas = dataA['aa'].values
aas = list(set(aas))
aas.sort()
georep = psu.GeoReport([],pdbDataPath, edDataPath, printPath, ed=False, dssp=False, includePdbs=False,keepDisordered=False)
for trio in trios:
if perAA:
for aa in aas:
dataCutA = dataA.query("aa == '" + aa + "'")
dataCutB = dataB.query("aa == '" + aa + "'")
dataCutC = dataC.query("aa == '" + aa + "'")
dataCutD = dataD.query("aa == '" + aa + "'")
if len(trio) == 3:
georep.addScatter(data=dataCutA, geoX=trio[0], geoY=trio[1], hue=trio[2], title=aa + ':' + trio[0] + ':'+ trio[1] , palette='jet', sort='NON')
georep.addScatter(data=dataCutB, geoX=trio[0], geoY=trio[1], hue=trio[2], title=aa + ':' + trio[0] + ':' + trio[1], palette='jet', sort='NON')
georep.addScatter(data=dataCutC, geoX=trio[0], geoY=trio[1], hue=trio[2],title=aa + ':' + trio[0] + ':' + trio[1], palette='jet', sort='NON')
georep.addScatter(data=dataCutD, geoX=trio[0], geoY=trio[1], hue=trio[2], title=aa + ':' + trio[0] + ':' + trio[1], palette='jet', sort='NON')
else:
georep.addHistogram(data=dataCutA, geoX=trio[0],title=fourSetNames[0] + ' ' + trio[0], hue='ID')
georep.addHistogram(data=dataCutB, geoX=trio[0],title=fourSetNames[1] + ' ' + trio[0], hue='ID')
georep.addHistogram(data=dataCutC, geoX=trio[0],title=fourSetNames[2] + ' ' + trio[0], hue='ID')
georep.addHistogram(data=dataCutD, geoX=trio[0],title=fourSetNames[3] + ' ' + trio[0], hue='ID')
else:
if len(trio) == 3:
georep.addScatter(data=dataA, geoX=trio[0], geoY=trio[1], hue=trio[2],title=trio[0] + '|' + trio[1]+ '|' + trio[2] + ' Unrestricted', palette='jet', sort='NON')
georep.addScatter(data=dataB, geoX=trio[0], geoY=trio[1], hue=trio[2], title=trio[0] + '|' + trio[1]+ '|' + trio[2] + ' Restricted', palette='jet', sort='NON')
georep.addScatter(data=dataC, geoX=trio[0], geoY=trio[1], hue=trio[2], title=trio[0] + '|' + trio[1]+ '|' + trio[2] + ' Restricted+cut', palette='jet', sort='NON')
georep.addScatter(data=dataD, geoX=trio[0], geoY=trio[1], hue=trio[2], title=trio[0] + '|' + trio[1]+ '|' + trio[2] + ' Adjusted', palette='jet', sort='NON')
else:
georep.addHistogram(data=dataA, geoX=trio[0], title=fourSetNames[0] + ' ' + trio[0], hue='ID')
georep.addHistogram(data=dataB, geoX=trio[0], title=fourSetNames[1] + ' ' + trio[0], hue='ID')
georep.addHistogram(data=dataC, geoX=trio[0], title=fourSetNames[2] + ' ' + trio[0], hue='ID')
georep.addHistogram(data=dataD, geoX=trio[0], title=fourSetNames[3] + ' ' + trio[0], hue='ID')
georep.printToHtml(title, 4, pdbSet + '_Defensible' + tag)
def compareSets(tag):
import matplotlib.pyplot as plt
plt.close('all')
plt.clf()
plt.cla()
pdbDataPath = help.rootPath + '/ProteinDataFiles/pdb_data/'
edDataPath = '/home/rachel/Documents/Bioinformatics/ProteinDataFiles/ccp4_data/'
loadPath = help.rootPath + '/BbkProject/PhDThesis/0.Papers/3.DefensibleGeometry/EvidencedSet/DataD/'
printPath = help.rootPath + '/BbkProject/PhDThesis/0.Papers/3.DefensibleGeometry/EvidencedSet/DataE/'
geos = ['N:CA', 'CA:C', 'C:O', 'C:N+1', 'TAU', 'C-1:N:CA', 'CA:C:N+1', 'CA:C:O', 'O:C:N+1', 'CA:C:N+1']
aas = ['ALL', 'ALA', 'CYS', 'ASP', 'GLU', 'PHE', 'GLY', 'HIS', 'ILE', 'LYS', 'LEU', 'MET', 'ASN', 'PRO', 'GLN','ARG', 'SER', 'THR', 'VAL', 'TRP', 'TYR']
fileName = tag + 'Data_SetsSummaryMerged.csv'
data = pd.read_csv(loadPath + fileName)
georep = psu.GeoReport([], pdbDataPath, edDataPath, printPath, ed=False, dssp=False, includePdbs=False, keepDisordered=False)
for geo in geos:
dataCut = data.query('geo == "' + geo + '"')
dataCutCount = dataCut.query('count > 0')
dataCutCount = dataCutCount.query('aa != "' + 'ALL' + '"')
dataCutAll = dataCut.query('aa == "' + 'ALL' + '"')
dataCutPRO = dataCut.query('aa == "' + 'PRO' + '"')
dataCutGLY = dataCut.query('aa == "' + 'GLY' + '"')
georep.addScatter(data=dataCutAll, geoX='mean', geoY='set', hue='sd', title=geo + ' ALL (exc gly/pro)', palette='jet', categorical=False,sort='NON')
georep.addScatter(data=dataCutGLY, geoX='mean', geoY='set', hue='sd', title=geo + ' GLY', palette='jet', categorical=False,sort='NON')
georep.addScatter(data=dataCutPRO, geoX='mean', geoY='set',hue='sd', title=geo + ' PRO', palette='jet', categorical=False,sort='NON')
georep.addScatter(data=dataCutCount, geoX='count', geoY='aa', hue='set', title=geo + ' Best Supported counts per aa', palette='jet_r', categorical=True, sort='NON')
georep.addScatter(data=dataCutCount, geoX='mean', geoY='aa', hue='set', title=geo + ' Best Supported means per aa', palette='jet_r',categorical=True, sort='NON')
georep.addScatter(data=dataCutCount, geoX='sd', geoY='aa', hue='set', title=geo + ' Best Supported sd per aa', palette='jet_r', categorical=True, sort='NON')
georep.printToHtml('Best Supported and Engh&Huber Compare', 3, tag + 'Compare_EH_Sets')
pdbDataPath = '/home/rachel/Documents/Bioinformatics/ProteinDataFiles/pdb_data/'
edDataPath = '/home/rachel/Documents/Bioinformatics/ProteinDataFiles/ccp4_data/'
printPath = '/home/rachel/Documents/Bioinformatics/ProteinDataFiles/results_psu/density/'
### split list in 2 for memory purposes
#pdbList = ['1ejg','1us0','1tt8','1i1w','1ucs','1yk4','1yk4','1hje','1r6j']
#pdbList = ['2bw4','3nir','3x2m','2VB1','3A39','2b97','2OV0','2WFI']
#pdbList = ['4ZM7','4REK','4ZM7','5D8V','5NW3','5qkw']
pdbList = [
'6jvv', '6rr2', '6E6O', '6S2M', '6shk', '6fgz', '6ctd', '6fwf', '6q53'
]
#pdbList = ['1ejg','1us0','1tt8','1i1w','1ucs','1yk4','1yk4','1hje','1r6j','2bw4','3nir','3x2m','2VB1','3A39','2b97','2OV0','2WFI','3o4p','1pjx']
#pdbList = ['4ZM7','4REK','4ZM7','5D8V','5NW3','5qkw','6jvv','6rr2','6E6O','6S2M','6shk','6fgz','6ctd','6fwf','6q53']
### split list in 2 for memory purposes
#peaksList=['1ejg','1us0','1tt8','1i1w','1ucs','6jvv','5nqo']
peaksList = [] #['1us0','1tt8']
#peaksList=['1i1w','1ucs','5nqo']
#pointsList=['6fwf','6q53','6ctd','6fgz','6rr2','6shk','6rr2']
pointsList = ['1ejg']
for pdb in peaksList:
georep = geor.GeoReport([pdb], pdbDataPath, edDataPath, printPath)
georep.printReport('Slow_DensityPeaksPerPdb', pdb + '_denpk')
for pdb in pointsList:
georep = geor.GeoReport([pdb], pdbDataPath, edDataPath, printPath)
georep.printReport('Slow_DensityPointsPerPdb', pdb + '_denpt')
dic = {}
A1 = randomOnSphere(A1_atom, 0.01) #7)
A2 = randomOnSphere(A2_atom, 0.01) #5)
A3 = randomOnSphere(A3_atom, 0.01) #7)
a1a2a3 = calcs.angle(A1[0], A1[1], A1[2], A2[0], A2[1], A2[2], A3[0],
A3[1], A3[2])
a1a2 = calcs.distance(A1[0], A1[1], A1[2], A2[0], A2[1], A2[2])
a2a3 = calcs.distance(A2[0], A2[1], A2[2], A3[0], A3[1], A3[2])
dic['pdbCode'] = 'Iter_' + str(count)
dic['chain'] = 'A'
dic['rid'] = 1
dic['ANGLE'] = a1a2a3
dic['A1:A2'] = a1a2
dic['A2:A3'] = a2a3
vals.append(dic)
dataFrame = pd.DataFrame.from_dict(vals)
georep = psu.GeoReport([],
pdbDataPath,
edDataPath,
printPath,
ed=False,
dssp=False,
includePdbs=False,
keepDisordered=False)
georep.addHistogram(data=dataFrame, geoX='ANGLE', title='')
georep.addHistogram(data=dataFrame, geoX='A1:A2', title='')
georep.addHistogram(data=dataFrame, geoX='A2:A3', title='')
georep.printToHtml('Simulated Atoms', 3, 'SimReport')
import pandas as pd
import Ch000_Functions as help
from PsuGeometry import GeoReport as psu
pdblist = help.getPDBList100()
pdblist.sort()
#pdblist = pdblist[:10]
hueList = ['aa', 'rid', 'bfactor', 'pdbCode', 'bfactorRatio', 'disordered','occupancy','dssp']
dsspPrintPath = '../../PdbLists/'
georep = psu.GeoReport(pdblist, help.pdbDataPathLx, help.edDataPath, dsspPrintPath, ed=False, dssp=True, includePdbs=False, keepDisordered=True)
datacsv = georep.getGeoemtryCsv(['N:CA'],hueList)
datacsv = datacsv[['pdbCode','chain','rid','aa','dssp']]
print(datacsv)
if False:#don;t accidentally run this and replace it
datacsv.to_csv(dsspPrintPath + 'dssp.csv', index=False)
print(datacsv)
pdbDataPath = '/home/rachel/Documents/Bioinformatics/ProteinDataFiles/pdb_data/'
edDataPath = '/home/rachel/Documents/Bioinformatics/ProteinDataFiles/ccp4_data/'
printPath = '/home/rachel/Documents/Bioinformatics/ProteinDataFiles/results_psu/bad/'
# Create the geoemtric data
geoPsi = ['N:O', 'CB:O', 'N:CA:C:N+1']
geoListMain = ['CA:C', 'N:CA', 'C:O']
hueList = ['dssp', 'aa', 'bfactor', '2FoFc',
'rid'] # note the hues are the sum od the atoms
pdbList = ['2lc9', '2lcb', '2cnq', '1i1w'] # structures with errors
for pdbCode in pdbList:
georep = psu.GeoReport([pdbCode], pdbDataPath, edDataPath, printPath)
dataPsi = georep.getGeoemtryCsv(geoPsi, hueList)
dataMain = georep.getGeoemtryCsv(geoListMain, hueList)
#Create the geoplots
printList = []
georep.addHistogram(geoX='N:CA', title='N-CA', ghost=True, hue='rid')
georep.addHistogram(geoX='CA:C', title='C-CA', ghost=True, hue='rid')
georep.addHistogram(geoX='CA:CA+1', title='CA-CA+1', ghost=True, hue='rid')
georep.addHistogram(data=dataMain,
geoX='N:CA',
title='N-CA',
ghost=True,
splitKey='pdbCode')
###################################################################################
pdbDataPath = '/home/rachel/Documents/Bioinformatics/ProteinDataFiles/pdb_data/'
edDataPath = '/home/rachel/Documents/Bioinformatics/ProteinDataFiles/ccp4_data/'
printPath = '/home/rachel/Documents/Bioinformatics/ProteinDataFiles/results_psu/Paper02/'
dsspHue='dssp'
includeDSSP = False
if myWindowsLaptop:
pdbDataPath = 'F:/Code/ProteinDataFiles/pdb_data/'
edDataPath = 'F:/Code/ProteinDataFiles/ccp4_data/'
printPath = 'F:/Code/ProteinDataFiles/results_psu/Paper02/'
includeDSSP = False # on my windows computer
###########################################################################################
georepData = psu.GeoReport(pdbList1000, pdbDataPath, edDataPath, printPath, ed=False, dssp=includeDSSP, includePdbs=True)
geoList = []
for geo in dihs:
geoList.append(geo)
for geo in distances:
geoList.append(geo)
for geo in angles:
geoList.append(geo)
count = 0
length = len(pdbList1000)
for pdb in pdbList1000:
print(pdb,' ', count,'/',length)
count += 1
def maximaCompareReal(pdbSet, pdbList, tag):
pdbDataPath = help.rootPath + '/ProteinDataFiles/pdb_out/' + pdbSet
edDataPath = '/home/rachel/Documents/Bioinformatics/ProteinDataFiles/ccp4_data/'
printPath = help.rootPath + '/BbkProject/PhDThesis/0.Papers/3.DefensibleGeometry/EvidencedSet/DataM/'
realCsv, badRealCsv, occRealCsv = help.getMaximaDiffs(
pdbSet, pdbList, False)
realOccOne = realCsv.query("Occupancy == 1")
realCutDown = realOccOne.query("BFactor < 10")
realCutDown5 = realCutDown.query("Difference <= 0.05")
realCol = 'brg' #'RdPu'
georep = psu.GeoReport([],
pdbDataPath,
edDataPath,
printPath,
ed=False,
dssp=False,
includePdbs=False,
keepDisordered=False)
georep.addHistogram(data=realCsv,
geoX='Difference',
title='PDB structures',
count=True,
hue='pdbCode')
georep.addHistogram(data=realCutDown,
geoX='Difference',
title='PDB structures, Occ=1, BFact<10',
count=True,
hue='pdbCode')
georep.addHistogram(
data=realCutDown5,
geoX='Difference',
title='PDB structures, Occ=1, BFact<10, Difference<=0.05',
count=True,
hue='pdbCode')
georep.addScatter(data=realCsv,
geoX='Difference',
geoY='BFactor',
hue='BGridDistance',
palette=realCol,
sort='RANDOM',
title='Occ=1')
georep.addScatter(data=realCutDown,
geoX='Difference',
geoY='BFactor',
hue='BGridDistance',
palette=realCol,
sort='RANDOM',
title='bfactor<=10')
georep.addScatter(data=realCutDown5,
geoX='Difference',
geoY='BFactor',
hue='BGridDistance',
palette=realCol,
sort='RANDOM',
title='Diff<=0.05')
georep.addScatter(data=realCsv,
geoX='Difference',
geoY='Width',
hue='BFactor',
categorical=False,
palette=realCol,
sort='RANDOM',
title='Occ=1')
georep.addScatter(data=realCutDown,
geoX='Difference',
geoY='Width',
hue='BFactor',
categorical=False,
palette=realCol,
sort='RANDOM',
title='BFactor<=10')
georep.addScatter(data=realCutDown5,
geoX='Difference',
geoY='Width',
hue='BFactor',
categorical=False,
palette=realCol,
sort='RANDOM',
title='Diif<=0.05')
#for pdb in pdbList:
# print(pdb)
# georep.addHistogram(data=realCsv, geoX='Difference', title='Occ=1', hue='AtomNo', count=True, restrictions={'pdbCode': pdb})
# georep.addHistogram(data=realCutDown, geoX='Difference', title='Occ=1, BFact<10', hue='AtomNo', count=True, restrictions={'pdbCode': pdb})
# georep.addScatter(data=realCutDown, geoX='Difference', geoY='BGridDistance', hue='GridDistance', categorical=False, restrictions={'pdbCode': pdb}, palette=realCol, sort='RANDOM', title='')
# georep.addScatter(data=realCutDown, geoX='Difference', geoY='AtomType', hue='AtomNo', categorical=False, restrictions={'pdbCode': pdb}, palette=realCol, sort='RANDOM', title='')
georep.printToHtml('Maxima differences in PDB Structures, set=' + pdbSet,
3, 'Maxima_' + pdbSet + tag)
pdbDataPath = '/home/rachel/Documents/Bioinformatics/ProteinDataFiles/pdb_data/'
edDataPath = '/home/rachel/Documents/Bioinformatics/ProteinDataFiles/ccp4_data/'
printPath = '/home/rachel/Documents/Bioinformatics/ProteinDataFiles/results_psu/'
from PsuGeometry import GeoReport as geor
# Create the GeoPdb object and the report object with just that single pdb
pdbs = ['2bw4','1ejg','1us0']
pdbs = ['1ejg']
georep = geor.GeoReport(pdbs,pdbDataPath,edDataPath,printPath,False,False)
# Choose the geometric calculations desired and the hues we might want to look at
title = 'Protein Warhol'
georep.addProbability(geoX='N:CA:C:CB',geoY='N:CA:C:N+1',title='', palette='Spectral',restrictions={'aa':'PRO,ALA'})
georep.addProbability(geoX='N:CA:C:CB',geoY='N:CA:C:N+1',title='', palette='twilight_shifted',restrictions={'aa':'ALA'})
georep.addProbability(geoX='N:CA:C:CB',geoY='N:CA:C:N+1',title='', palette='inferno',restrictions={'aa':'PRO'})
georep.addProbability(geoX='N:CA:C:CB',geoY='N:CA:C:N+1',title='', palette='viridis_r')
# And finally create the reort with a file name of choice
georep.printToHtml(title,2,'warhol')
# Choose the geometric calculations desired and the hues we might want to look at
'''
title = 'Protein Halo'
georep.addProbability(geoX='N:O',geoY='CB:O',title='', palette='Spectral')
georep.addProbability(geoX='N:O',geoY='CB:O',title='', palette='twilight_shifted')
georep.addProbability(geoX='N:O',geoY='CB:O',title='', palette='inferno')
georep.addProbability(geoX='N:O',geoY='CB:O',title='', palette='nipy_spectral_r')
# And finally create the reort with a file name of choice
georep.printToHtml(title,2,'angel')
'''
def createBadDensitySlices(pdbSet, atomCe, atomLi, atomPl):
import matplotlib.pyplot as plt
plt.close('all')
plt.clf()
plt.cla()
pdbOriginalPath = help.rootPath + '/ProteinDataFiles/pdb_data/'
pdbDataPath = help.rootPath + '/ProteinDataFiles/pdb_out/' + pdbSet + '/'
edDataPath = help.rootPath + '/ProteinDataFiles/ccp4_data/'
printPath = help.rootPath + '/BbkProject/PhDThesis/0.Papers/3.DefensibleGeometry/EvidencedSet/SlicesF/'
# This gets the list of pdbs
pdbdata = pd.read_csv(
'../../PdbLists/Pdbs_Evidenced.csv'
) # This is a list of pdbs <= 1.1A non homologous to 90%
pdbListIn = pdbdata['PDB'].tolist()[0:]
#pdbListIn = ['1p1x']
for pdb in pdbListIn:
slicesList = []
#fileName = (pdbDataPath + 'pdb' + pdb + '_' + atomCe + atomLi + atomPl + '.bad').lower()
fileNameIn = (pdbDataPath + 'pdb' + pdb + '.bad').lower()
print(fileNameIn)
import os.path
if os.path.isfile(fileNameIn):
text_file = open(fileNameIn, "r")
lines = text_file.read().split('\n')
print(len(lines))
text_file.close()
for line in lines:
atom = line[12:14].lstrip().rstrip()
aa = line[14:20].lstrip().rstrip()
chain = line[20:22].lstrip().rstrip()
rid = line[22:27].lstrip().rstrip()
print(pdb, atom, aa, chain, rid, line)
if rid != '':
if [pdb, chain, rid] not in slicesList:
slicesList.append([pdb, chain, rid])
bigstring = ""
for sl in slicesList:
print(sl)
georep = psu.GeoReport([sl[0]],
pdbOriginalPath,
edDataPath,
printPath,
ed=False,
dssp=False)
pdbmanager = geopdb.GeoPdbs(pdbOriginalPath,
edDataPath,
ed=False,
dssp=False)
apdb = pdbmanager.getPdb(sl[0], True)
pdbcsv = apdb.getDataFrame()
queryC = 'rid==' + str(
sl[2]
) + ' and chain=="' + sl[1] + '"' + ' and atom=="' + atomCe + '"'
queryL = 'rid==' + str(
sl[2]
) + ' and chain=="' + sl[1] + '"' + ' and atom=="' + atomLi + '"'
queryP = 'rid==' + str(
sl[2]
) + ' and chain=="' + sl[1] + '"' + ' and atom=="' + atomPl + '"'
dataC = pdbcsv.query(queryC)
dataL = pdbcsv.query(queryL)
dataP = pdbcsv.query(queryP)
if len(dataC) > 0 and len(dataL) > 0 and len(dataP) > 0:
cx = round(dataC['x'].values[0], 3)
cy = round(dataC['y'].values[0], 3)
cz = round(dataC['z'].values[0], 3)
lx = round(dataL['x'].values[0], 3)
ly = round(dataL['y'].values[0], 3)
lz = round(dataL['z'].values[0], 3)
px = round(dataP['x'].values[0], 3)
py = round(dataP['y'].values[0], 3)
pz = round(dataP['z'].values[0], 3)
row = sl[0] + "," + sl[1] + str(
sl[2]) + "," + str(cx) + "," + str(cy) + "," + str(cz)
row += "," + str(lx) + "," + str(ly) + "," + str(lz)
row += "," + str(px) + "," + str(py) + "," + str(pz)
print(row)
bigstring += row + '\n'
if len(slicesList) > 0:
print("########RESULTS#########")
print("")
print(bigstring)
tag = atomCe + atomLi + atomPl
f = open(
printPath + 'BadSlice_' + pdbSet + '_' + pdb + '_' + tag +
'.txt', "w")
f.write(bigstring)
f.close()
edSlicePath = 'F:/Code/ProteinDataFiles/ccp4_out/'
#printPath = 'F:/Code/ProteinDataFiles/results_psu/Paper02/'
#We are going to load the data that has been created by density flight
edSlicePath += FileDir + "/"
print(edSlicePath + "_Results.csv")
inputdata = pd.read_csv(edSlicePath + "_Results.csv")
#PdbCode,Tag,
# CentreX,CentreY,CentreZ,# LinearX,LinearY,LinearZ,PlanarX,PlanarY,PlanarZ,
# CentreV,LinearV,PlanarV,Angle,
# BCentreX,BCentreY,BCentreZ,BLinearX,BLinearY,BLinearZ,BPlanarX,BPlanarY,BPlanarZ,
# BCentreC,BLinearV,BPlanarV,BAngle
georep = psu.GeoReport([],
pdbDataPath,
edDataPath,
edSlicePath,
ed=False,
dssp=False)
#georepPrint = psu.GeoReport([],pdbDataPath,edDataPath,edSlicePath,ed=False,dssp=False)
pdbs = inputdata['PdbCode'].values
tags = inputdata['Tag'].values
taus = inputdata['Angle'].values
btaus = inputdata['BAngle'].values
origs = []
betters = []
radiants = []
brads = []
# Once the app has created the data we can load it
It runs correlation reports on proline, which it colours on the hue of CHI1 and CA-1:CA
These geometric measures are proxies for up/down pucker of the proline run (up-pucker=-ve CHI1)
And cis-trans proline, where pre-omega means proline, which corresponds to short CA-1:CA
'''
pdbDataPath = '/home/rachel/Documents/Bioinformatics/ProteinDataFiles/pdb_data/'
edDataPath = '/home/rachel/Documents/Bioinformatics/ProteinDataFiles/ccp4_data/'
printPath = '/home/rachel/Documents/Bioinformatics/ProteinDataFiles/results_psu/1000Structures/'
pdbList = []
pdbdata = pd.read_csv('structures09.csv')
pdbList = pdbdata['pdb_code']
georep = geor.GeoReport(pdbList,
pdbDataPath,
edDataPath,
printPath,
ed=False,
dssp=False)
geoList = [
'PHI', 'PSI', 'TAU', 'C-1:C', 'C-1:N:CA', 'CHI1', 'CA-1:CA', 'OMEGA',
'CA:CA+1', 'C-1:N:CA:C'
]
hueList = ['aa', 'bfactor', 'rid', 'resolution', 'pdbCode']
data = georep.getGeoemtryCsv(geoList, hueList)
georep.addScatter(data=data,
geoX='PHI',
geoY='PSI',
hue='CHI1',
def makeSlicesHtml(setName, fileNameOrig,fileNameAdj, title,titleType,tag):
import matplotlib.pyplot as plt
plt.close('all')
plt.clf()
plt.cla()
FileDir = setName
firstRow = 1
outfileName = titleType + '_' + setName + "_" + tag
pdbDataPath = help.rootPath + '/ProteinDataFiles/pdb_data/'
edDataPath = help.rootPath + '/ProteinDataFiles/ccp4_data/'
edSlicePath = help.rootPath + '/ProteinDataFiles/ccp4_out/'
printPath = help.rootPath + '/BbkProject/PhDThesis/0.Papers/3.DefensibleGeometry/EvidencedSet/SlicesH/'
#We are going to load the data that has been created by density flight
edSlicePath += FileDir + "/"
print(edSlicePath + "_Results.csv")
inputdata = pd.read_csv(edSlicePath + "_Results.csv")
pdbs = inputdata['PdbCode'].values
tags = inputdata['Tag'].values
taus = inputdata['Angle'].values
valsPath = help.rootPath + '/BbkProject/PhDThesis/0.Papers/3.DefensibleGeometry/EvidencedSet/SlicesC/'
print(valsPath)
inputValsOrig = pd.read_csv(valsPath + fileNameOrig)
inputValsAdj = pd.read_csv(valsPath + fileNameAdj)
#inputVals = pd.read_csv(valsPath + "GoodOutliers_" + tag + ".csv")
#print(inputVals)
georep = psu.GeoReport([], pdbDataPath, edDataPath, printPath, ed=False, dssp=False)
#origs = []
#radiants = []
for i in range(0,len(pdbs)):
pdb = pdbs[i]
tag = tags[i]
tau = taus[i]
pdbInputValsOrig = inputValsOrig.query("pdbCode == '" + pdb + "'")
pdbInputValsAdj = inputValsAdj.query("pdbCode == '" + pdb + "'")
valsOrig = pdbInputValsOrig['value'].values
valsAdj = pdbInputValsAdj['value'].values
vpdbs = pdbInputValsAdj['pdbCode'].values
vaas = pdbInputValsAdj['aa'].values
rids = pdbInputValsAdj['rid'].values
chains = pdbInputValsAdj['chain'].values
for j in range(0,len(vpdbs)):
vpdb = vpdbs[j]
newTag = vaas[j] + chains[j] + str(rids[j])
print(vpdb,pdb,tag,newTag)
if pdb == vpdb and newTag in tag:
sliceOrigVal = georep.loadSlice(edSlicePath + pdb + tag + "value_slice.csv")
sliceOrigRad = georep.loadSlice(edSlicePath + pdb + tag + "radiant_slice.csv")
sliceOrigMag = georep.loadSlice(edSlicePath + pdb + tag + "magnitude_slice.csv")
'''
https://stackoverflow.com/questions/16400241/how-to-redefine-a-color-for-a-specific-value-in-a-matplotlib-colormap/16401183#16401183
'''
sliceOrigPos = georep.loadSlice(edSlicePath + pdb + tag + "poses_slice.csv")
#This takes the data from the electron density only
imtitle = pdb +' ' + tag + ' value, angle=' + str(round(tau,3))
# This takes the data from seperately created outliers file
if '_O' in tag:
imtitle = pdb + ' ' + tag + ' value=' + str(round(valsOrig[j], 3))
else:
imtitle = pdb + ' ' + tag + ' value=' + str(round(valsAdj[j], 3))
#if pdb != vpdb:
# print(pdb,vpdb)
# imtitle = title = pdb +' ' + tag + ' value, angle=' + str(round(tau,3)) + ' Error loading outlier'
georep.addSlice(sliceOrigVal, palette='cubehelix_r', title=imtitle, YellowDots=sliceOrigPos,Contour=True)
georep.addSlice(sliceOrigRad, palette='bone',title=pdb + ' ' + tag + ' radiant',Contour=False,YellowDots=sliceOrigPos)
georep.addSlice(sliceOrigMag, palette='bone', title=pdb + ' ' + tag + ' magnitude', Contour=True,YellowDots=sliceOrigPos)
#origs.append(sliceOrigVal)
#radiants.append(sliceOrigRad)
firstRow += 1
#georep.addSlices(origs, palette='cubehelix_r', title='Average values', logged=False, centre=False)
#georep.addSlices(radiants, palette='bone', title='Average radiant', logged=False, centre=False,Contour=False)
georep.printToHtml(title,6,outfileName)
'''
This script looks electron density correlations
'''
pdbDataPath = '/home/rachel/Documents/Bioinformatics/ProteinDataFiles/pdb_data/'
edDataPath = '/home/rachel/Documents/Bioinformatics/ProteinDataFiles/ccp4_data/'
printPath = '/home/rachel/Documents/Bioinformatics/ProteinDataFiles/results_psu/Paper01/'
pdbList = ['2bw4','5nqo','1ejg','6q53']
geoList = ['PHI','PSI','TAU']
hueList = ['aa','bfactor','rid','resolution','pdbCode','2FoFc','dssp']
georep2 = psu.GeoReport(pdbList, pdbDataPath, edDataPath, printPath)
for pdb in pdbList:
georep = psu.GeoReport([pdb], pdbDataPath, edDataPath, printPath)
#georep.printReport('DataPerPdb', 'Results9_data')
#georep.printReport('Slow_DensityPeaksPerPdb', 'Results9_density')
data = georep.getGeoemtryCsv(geoList, hueList)
georep2.addScatter(data=data, geoX='ridx', geoY='2FoFc', hue='TAU', title=pdb + ' Backbone tau', palette='jet',sort='NON', vmin=106, vmax=116)
georep2.addScatter(data=data, geoX='ridx', geoY='2FoFc', hue='PHI', title=pdb + ' Backbone phi', palette='jet',sort='NON', vmin=-170, vmax=170)
georep2.addScatter(data=data, geoX='ridx', geoY='2FoFc', hue='PSI', title=pdb + ' Backbone psi', palette='jet',sort='NON', vmin=-170, vmax=170)
georep2.addScatter(data=data, geoX='ridx', geoY='2FoFc', hue='dssp', title=pdb + ' Backbone dssp', palette='jet_r', sort='NON')
mergedDataSet['CID'] = mergedDataSet['pdbCode'] + mergedDataSet[
'chain'] + mergedDataSet['rid']
mergedDataSet = mergedDataSet.set_index('CID').join(
ccContacts.set_index('CID'))
mergedDataSet.to_csv(help.loadPath + "MergedWithContacts.csv", index=False)
mergedDataSet = mergedDataSet.dropna()
#qu = "dssp == 'E' or dssp == 'B' or dssp == '-' "
#qu = qu + "or dssp == 'T' or dssp == 'S' or dssp == 'H' or dssp == 'G' or dssp == 'I')"
#mergedDataSet = mergedDataSet.query(qu)
# create a report based on contacts
georep = psu.GeoReport([],
"",
"",
help.printPath,
ed=False,
dssp=False,
includePdbs=False,
keepDisordered=False)
print('### Creating reports ###')
georep.addScatter(data=mergedDataSet,
geoX='N:CA_Orig',
geoY='Contacts',
title='N:CA Original',
hue='dssp',
categorical=True,
sort='NON',
palette='jet_r')
georep.addScatter(data=mergedDataSet,
geoX='N:CA_Diff',
def maximaCompareFake(pdbSet, pdbList, tag, reduce):
print('Plotting fake maxima differences', pdbSet)
pdbDataPath = help.rootPath + '/ProteinDataFiles/pdb_out/' + pdbSet
edDataPath = '/home/rachel/Documents/Bioinformatics/ProteinDataFiles/ccp4_data/'
printPath = help.rootPath + '/BbkProject/PhDThesis/0.Papers/3.DefensibleGeometry/EvidencedSet/DataM/'
realCsv, badRealCsv, fakeCsv, badFakeCsv = help.getMaximaDiffs(
pdbSet, pdbList, True)
if reduce:
#fakeCsv1 = fakeCsv.query('BGridDistance < 1.7')
#fakeCsv2 = fakeCsv1.query('BGridDistance < 0.95')
#fakeCsv1 = fakeCsv1.query('BGridDistance > 1.05')
#fakeCsv3 = fakeCsv1.query('BGridDistance < 1.35')
#fakeCsv1 = fakeCsv1.query('BGridDistance > 1.45')
#fakeCsv = pd.concat([fakeCsv1,fakeCsv2,fakeCsv3])
fakeCsv = fakeCsv.query('Difference <= 0.05')
#fakeCsv2 = fakeCsv1.query('BGridDistance < 1.4 or BGridDistance > 1.42')
#fakeCsv3 = fakeCsv2.query('BGridDistance < 0.95 or BGridDistance > 1.05')
print(fakeCsv)
fakeCol = 'jet_r' #'GnBu'
georep = psu.GeoReport([],
pdbDataPath,
edDataPath,
printPath,
ed=False,
dssp=False,
includePdbs=False,
keepDisordered=False)
georep.addHistogram(data=fakeCsv,
geoX='Difference',
title='Fake PDB structures, Occ=1, BFact=2',
count=True,
hue='pdbCode',
palette='LightSeaGreen')
georep.addScatter(data=fakeCsv,
geoX='Difference',
geoY='pdbCode',
hue='Width',
categorical=False,
palette=fakeCol + '',
sort='RANDOM',
title='')
georep.addScatter(data=fakeCsv,
geoX='Difference',
geoY='Width',
hue='pdbCode',
categorical=True,
palette='tab20',
sort='RANDOM',
title='')
georep.addScatter(data=fakeCsv,
geoX='Difference',
geoY='GridDistance',
hue='BGridDistance',
categorical=False,
palette=fakeCol,
sort='RANDOM',
title='')
georep.addScatter(data=fakeCsv,
geoX='Difference',
geoY='BGridDistance',
hue='GridDistance',
categorical=False,
palette=fakeCol,
sort='RANDOM',
title='')
georep.addScatter(data=fakeCsv,
geoX='GridDistance',
geoY='BGridDistance',
hue='Difference',
categorical=False,
palette=fakeCol,
sort='RANDOM',
title='')
georep.addProbability(data=fakeCsv,
geoX='Difference',
geoY='GridDistance',
palette='cubehelix_r')
georep.addProbability(data=fakeCsv,
geoX='Difference',
geoY='BGridDistance',
palette='cubehelix_r')
georep.addProbability(data=fakeCsv,
geoX='GridDistance',
geoY='BGridDistance',
palette='cubehelix_r')
for pdb in pdbList:
print(pdb)
georep.addHistogram(data=fakeCsv,
geoX='Difference',
title='Fake Density, Occ=1, BFact=10',
hue='AtomNo',
count=True,
restrictions={'pdbCode': pdb},
palette='LightSeaGreen')
georep.addHistogram(data=fakeCsv,
geoX='Difference',
title='Fake Density, Occ=1, BFact=10',
hue='Reason',
count=True,
restrictions={'pdbCode': pdb},
palette='LightSeaGreen')
georep.addScatter(data=fakeCsv,
geoX='AtomNo',
geoY='Difference',
hue='AtomType',
categorical=True,
restrictions={'pdbCode': pdb},
palette='tab20',
sort='RANDOM',
title='')
georep.printToHtml('Maxima differences, set=' + pdbSet, 3,
'Maxima_' + pdbSet + tag)
hueList = ['dssp', 'aa', 'rid', 'bfactor']
aas = ['GLY']
###################################################################################
pdbDataPath = '/home/rachel/Documents/Bioinformatics/ProteinDataFiles/pdb_data/'
edDataPath = '/home/rachel/Documents/Bioinformatics/ProteinDataFiles/ccp4_data/'
printPath = '/home/rachel/Documents/Bioinformatics/ProteinDataFiles/results_psu/Paper02/'
if myWindowsLaptop:
pdbDataPath = 'F:/Code/ProteinDataFiles/pdb_data/'
edDataPath = 'F:/Code/ProteinDataFiles/ccp4_data/'
printPath = 'F:/Code/ProteinDataFiles/results_psu/Paper02/'
###########################################################################################
georep = psu.GeoReport(pdbList1000,
pdbDataPath,
edDataPath,
printPath,
ed=False,
dssp=False,
keepDisordered=keepDisordered,
includePdbs=False)
data = georep.getGeoemtryCsv(geoList, hueList, bfactorFactor)
#data = data.query('TAU > 100')
#data = data.query('TAU < 125')
dataPsiRange = data.query('PSI > -50')
dataPsiRange = dataPsiRange.query('PSI < 50')
for aa in aas:
sql = 'aa == "' + aa + '"'
dataaa = data.query(sql)
dataPsiRangeaa = dataPsiRange.query(sql)
georep.addScatter(data=dataaa,
def compareAtomsPdbAdjusted(dataCombined, geos, pdbSet, tag):
pdbDataPath = help.rootPath + '/ProteinDataFiles/pdb_out/' + pdbSet
edDataPath = '/home/rachel/Documents/Bioinformatics/ProteinDataFiles/ccp4_data/'
printPath = help.rootPath + '/BbkProject/PhDThesis/0.Papers/3.DefensibleGeometry/EvidencedSet/DataM/'
georep = psu.GeoReport([],
pdbDataPath,
edDataPath,
printPath,
ed=False,
dssp=False,
includePdbs=False,
keepDisordered=False)
for geo in geos:
dataCombined[
geo +
'_Diff'] = dataCombined[geo + '_Orig'] - dataCombined[geo + '_Adj']
georep.addHistogram(data=dataCombined,
geoX=geo + '_Orig',
title='Pdb Atoms',
count=True,
hue='pdbCode')
georep.addHistogram(data=dataCombined,
geoX=geo + '_Adj',
title='Adjusted Atoms',
count=True,
hue='pdbCode')
#georep.addScatter(data=dataCombined, geoX=geo + '_Diff', geoY='RES', hue='SOFTWARE', palette='jet_r', sort='RANDOM', categorical=True, title='Resolution and atom differences ' + geo)
#georep.addScatter(data=dataCombined, geoX=geo + '_Diff', geoY='SOFTWARE', hue='RES', palette='viridis_r', sort='DESC', categorical=False, title='Resolution and atom differences ' + geo)
georep.addHexBins(data=dataCombined,
geoX=geo + '_Diff',
geoY='RES',
title='Count ' + geo,
hue='count',
palette='cubehelix_r')
georep.addScatter(data=dataCombined,
geoX=geo + '_Orig',
geoY=geo + '_Adj',
hue='SOFTWARE',
palette='jet_r',
sort='RANDOM',
categorical=True,
title='Software and atom positions ' + geo)
georep.addScatter(data=dataCombined,
geoX=geo + '_Orig',
geoY=geo + '_Adj',
hue='RES',
palette='viridis_r',
sort='DESC',
categorical=False,
title='Resolution and atom positions ' + geo)
georep.addHexBins(data=dataCombined,
geoX=geo + '_Orig',
geoY=geo + '_Adj',
title='Count ' + geo,
hue='count',
palette='cubehelix_r')
#georep.addScatter(data=dataCombined, geoX=geo + '_Orig', geoY='SOFTWARE', hue=geo + '_Adj', palette='jet_r', sort='RANDOM', categorical=True, title='Comparing atom positions ' + geo)
#for pdb in pdbList:
# print(pdb)
# georep.addHistogram(data=realCsv, geoX='Difference', title='Occ=1', hue='AtomNo', count=True, restrictions={'pdbCode': pdb})
# georep.addHistogram(data=realCutDown, geoX='Difference', title='Occ=1, BFact<10', hue='AtomNo', count=True, restrictions={'pdbCode': pdb})
# georep.addScatter(data=realCutDown, geoX='Difference', geoY='BGridDistance', hue='GridDistance', categorical=False, restrictions={'pdbCode': pdb}, palette=realCol, sort='RANDOM', title='')
# georep.addScatter(data=realCutDown, geoX='Difference', geoY='AtomType', hue='AtomNo', categorical=False, restrictions={'pdbCode': pdb}, palette=realCol, sort='RANDOM', title='')
georep.printToHtml(
'Comparing atom positions: PDB vs maxima, set=' + pdbSet, 3,
'Compare_' + pdbSet + tag)
def getCsv(pdbSet,
pdbListIn,
geos,
badAtoms,
reloadPdb,
reloadCsv,
aa='ALL',
includeCis=False,
allAtoms=False,
bFactorFactor=1.3,
cutoff=0):
print('Getting CSV for', pdbSet)
pdbDataPath = rootPath + '/ProteinDataFiles/pdb_out/' + pdbSet + '/'
if pdbSet == 'PDB':
pdbDataPath = rootPath + '/ProteinDataFiles/pdb_data/'
edDataPath = '/home/rachel/Documents/Bioinformatics/ProteinDataFiles/ccp4_data/'
loadPath = rootPath + '/BbkProject/PhDThesis/0.Papers/3.DefensibleGeometry/EvidencedSet/DataB/'
printPath = rootPath + '/BbkProject/PhDThesis/0.Papers/3.DefensibleGeometry/EvidencedSet/DataK/'
fileName = 'Data_DefensibleWithGeosALL_' + pdbSet + '.csv'
if reloadCsv:
from PsuGeometry import GeoPdb as geopdb
pdbmanager = geopdb.GeoPdbs(pdbDataPath, edDataPath, False, False,
False, badAtoms)
if reloadPdb:
pdbmanager.clear()
pdbmanager = geopdb.GeoPdbs(pdbDataPath, edDataPath, False, False,
False, badAtoms)
#pdbdata = pd.read_csv('../../PdbLists/Pdbs_Evidenced.csv') # This is a list of pdbs <= 1.1A non homologous to 90%
#pdbListIn = pdbdata['PDB'].tolist()[0:]
#if cutoff > 0:
# pdbListIn = pdbdata['PDB'].tolist()[0:cutoff]
pdbList = []
for pdb in pdbListIn:
import os.path
filePdb = pdbDataPath + 'pdb' + pdb + '.ent'
#print('- Adding to csv',filePdb)
if os.path.isfile((filePdb).lower()):
pdbList.append(pdb.lower())
else:
print('No file:', pdbDataPath, pdb)
pdbList.sort()
hueList = [
'aa', 'rid', 'bfactor', 'pdbCode', 'bfactorRatio', 'disordered'
]
georep = psu.GeoReport(pdbList,
pdbDataPath,
edDataPath,
printPath,
ed=False,
dssp=False,
includePdbs=False,
keepDisordered=allAtoms)
if includeCis:
geos.append('CA-1:C-1:N:CA')
print('geoList', geos)
dataBest = georep.getGeoemtryCsv(geos,
hueList,
bFactorFactor,
allAtoms=allAtoms,
restrictedAa=aa)
try:
dataBest['rid'] = dataBest['rid'].astype(str)
dataBest['ID'] = dataBest['pdbCode'] + dataBest[
'chain'] + dataBest['rid'] + dataBest['aa']
except:
print('empty csv')
else:
dataBest = pd.read_csv(loadPath + fileName)
#aas = ['ALA', 'CYS', 'ASP', 'GLU', 'PHE', 'GLY', 'HIS', 'ILE', 'LYS', 'LEU', 'MET', 'ASN', 'PRO', 'GLN', 'ARG','SER', 'THR', 'VAL', 'TRP', 'TYR']
if includeCis:
dataBest['aa'] = dataBest.apply(
lambda row: applyCis(row['aa'], row['CA-1:C-1:N:CA']), axis=1)
if aa != 'ALL':
dataBest = dataBest.query('aa == "' + aa + '"')
return dataBest
includeDSSP = True
###################################################################################
pdbDataPath = '/home/rachel/Documents/Bioinformatics/ProteinDataFiles/pdb_data/'
edDataPath = '/home/rachel/Documents/Bioinformatics/ProteinDataFiles/ccp4_data/'
printPath = '/home/rachel/Documents/Bioinformatics/ProteinDataFiles/results_psu/Paper02/'
if myWindowsLaptop:
pdbDataPath = 'F:/Code/ProteinDataFiles/pdb_data/'
edDataPath = 'F:/Code/ProteinDataFiles/ccp4_data/'
printPath = 'F:/Code/ProteinDataFiles/results_psu/Paper02/'
includeDSSP = False # on my windows computer
###########################################################################################
georep = psu.GeoReport(pdbList1000,
pdbDataPath,
edDataPath,
printPath,
ed=False,
dssp=includeDSSP,
includePdbs=False)
data = georep.getGeoemtryCsv(geoList, hueList)
#datacorr = data.corr()
#sns.heatmap(datacorr, annot=True, cmap="vlag", vmin=-1, vmax=1)
#plt.show()
#Clean the data
data = data.drop('pdbCode', axis=1)
data = data.drop('chain', axis=1)
data = data.drop('rid', axis=1)
data = data.drop('aa', axis=1)
data = data.drop('ridx', axis=1)
data = data.drop('atomNo', axis=1)
geoLists.append(['5HB', ['N:O-2','C:O-2','N:CA:C:O-2','N:CA:N+1:O-2']])
# Hydrogen bond distances and dihedrals nearest O
geoLists.append(['6HBO', ['N:{O}','C:{O}','N:CA:C:{O}','N:CA:N+1:{O}']])
# Water
geoLists.append(['7WAT', ['N:HOH','C:HOH','N:CA:C:HOH','N:CA:N+1:HOH']])
# Other!
geoLists.append(['8XTRA', ['N:HETATM']])
hueList = ['aa', 'rid', 'bfactor','pdbCode','bfactorRatio','disordered','dssp']
aas = ['ALL']
print('Creating CSV files anew')
for geoListT in geoLists:
geoList = geoListT[1]
set = geoListT[0]
for aa in aas:
tag = 'Set' + set + aa
georep = psu.GeoReport(pdbList, pdbDataPath, edDataPath, printPath, ed=False, dssp=includeDSSP, includePdbs=False,keepDisordered=True)
print('Create unrestricted csv', geoList)
dataUnrestricted = georep.getGeoemtryCsv(geoList, hueList, -1,allAtoms=True,restrictedAa=aa)
dataUnrestricted.to_csv(printPath + 'CsvGeos_' + tag + '.csv', index=False)
print('----------Finished----------')
endx = time.time()
time_diff = endx - startx
timestring = str(int(time_diff / 60)) + "m " + str(int(time_diff % 60)) + "s"
print(timestring)
def EHCompare(pdbSet):
import matplotlib.pyplot as plt
plt.close('all')
plt.clf()
plt.cla()
pdbDataPath = help.rootPath + '/ProteinDataFiles/pdb_out/' + pdbSet + '/'
edDataPath = '/home/rachel/Documents/Bioinformatics/ProteinDataFiles/ccp4_data/'
loadPathEH = help.rootPath + '/BbkProject/PhDThesis/0.Papers/3.DefensibleGeometry/EvidencedSet/Data/'
loadPathCsv = help.rootPath + '/BbkProject/PhDThesis/0.Papers/3.DefensibleGeometry/EvidencedSet/DataB/'
printPath = help.rootPath + '/BbkProject/PhDThesis/0.Papers/3.DefensibleGeometry/EvidencedSet/DataC/'
#EH_SET,aa,N:CA,N:CA_SD,CA:C,CA:C_SD,C:O,C:O_SD,C:N+1,C:N+1_SD,N:CA:C,N:CA:C_SD,CA:C:N+1,CA:C:N+1_SD,CA:C:O,CA:C:O_SD,O:C:N+1,O:C:N+1_SD,C-1:N:CA,C-1:N:CA_SD
EHFileName = 'Data_EH.csv'
BestFileName = 'Data_DefensibleWithGeosALL_' + pdbSet + '.csv'
dataEH = pd.read_csv(loadPathEH + EHFileName)
dataBest = pd.read_csv(loadPathCsv + BestFileName)
aas = [
'ALA', 'CYS', 'ASP', 'GLU', 'PHE', 'GLY', 'HIS', 'ILE', 'LYS', 'LEU',
'MET', 'ASN', 'PRO', 'GLN', 'ARG', 'SER', 'THR', 'VAL', 'TRP', 'TYR'
]
geos = [
'N:CA', 'CA:C', 'C:O', 'C:N+1', 'TAU', 'CA:C:N+1', 'CA:C:O', 'O:C:N+1',
'C-1:N:CA'
]
#geos = ['N:CA','CA:C','C:O','C:N+1','N:CA:C']
#specifically looking at the mean and sd of the parameters in comparison to EH
print(dataEH)
georepAA = psu.GeoReport([],
pdbDataPath,
edDataPath,
printPath,
ed=False,
dssp=False,
includePdbs=False,
keepDisordered=False)
georepSummary = psu.GeoReport([],
pdbDataPath,
edDataPath,
printPath,
ed=False,
dssp=False,
includePdbs=False,
keepDisordered=False)
for geo in geos:
#Cut on ALL PRO and GLY
compareSets = ['1991', '2001']
listCompares = []
ehALL = dataEH.query("aa == 'ALL'")
ehGLY = dataEH.query("aa == 'GLY'")
ehPRO = dataEH.query("aa == 'PRO'")
ehCIS = dataEH.query("aa == 'CIS'")
bestALLCut = dataBest.query("aa != 'GLY'")
bestALLCut = bestALLCut.query("aa != 'PRO'")
bestGLYCut = dataBest.query("aa == 'GLY'")
bestPROCut = dataBest.query("aa == 'PRO'")
print(pdbSet, bestPROCut)
bestPROCut['ABSOMEGA'] = abs(bestPROCut['CA-1:C-1:N:CA'])
bestPROCis = bestPROCut.query("ABSOMEGA < 120")
bestPROTrans = bestPROCut.query("ABSOMEGA >= 120")
titleALL = ''
titleGLY = ''
titlePRO = ''
titleCIS = ''
for comp in compareSets:
ehSetALL = ehALL.query("EH_SET == '" + comp + "'")
ehSetGLY = ehGLY.query("EH_SET == '" + comp + "'")
ehSetPRO = ehPRO.query("EH_SET == '" + comp + "'")
ehSetCIS = ehCIS.query("EH_SET == '" + comp + "'")
meanALL = round(ehSetALL[geo].values[0], 3)
sdALL = ehSetALL[geo + '_SD'].values[0]
meanGLY = round(ehSetGLY[geo].values[0], 3)
sdGLY = ehSetGLY[geo + '_SD'].values[0]
meanPRO = round(ehSetPRO[geo].values[0], 3)
sdPRO = ehSetPRO[geo + '_SD'].values[0]
meanCIS = round(ehSetCIS[geo].values[0], 3)
sdCIS = ehSetCIS[geo + '_SD'].values[0]
titleALL += 'ALL ' + geo + ' ' + comp + ' Mean=' + str(
meanALL) + ' (' + str(sdALL) + ')\n'
titleGLY += 'GLY ' + geo + ' ' + comp + ' Mean=' + str(
meanGLY) + ' (' + str(sdGLY) + ')\n'
titlePRO += 'PRO ' + geo + ' ' + comp + ' Mean=' + str(
meanPRO) + ' (' + str(sdPRO) + ')\n'
titleCIS += 'CIS ' + geo + ' ' + comp + ' Mean=' + str(
meanCIS) + ' (' + str(sdCIS) + ')\n'
print(titleGLY, titlePRO, titleALL, titleCIS)
if geo == 'N:CA:C':
georepSummary.addHistogram(data=bestALLCut,
geoX='TAU',
title=titleALL)
georepSummary.addHistogram(data=bestGLYCut,
geoX='TAU',
title=titleGLY)
georepSummary.addHistogram(data=bestPROTrans,
geoX='TAU',
title=titlePRO)
georepSummary.addHistogram(data=bestPROCis,
geoX='TAU',
title=titleCIS)
else:
georepSummary.addHistogram(data=bestALLCut,
geoX=geo,
title=titleALL)
georepSummary.addHistogram(data=bestGLYCut,
geoX=geo,
title=titleGLY)
georepSummary.addHistogram(data=bestPROTrans,
geoX=geo,
title=titlePRO)
georepSummary.addHistogram(data=bestPROCis,
geoX=geo,
title=titleCIS)
'''
for aa in aas:
#prepare E&H comparison values
useaa = 'ALL'
if aa == 'PRO' or aa == 'GLY':
useaa = aa
ehCut = dataEH.query("aa == '" + useaa + "'")
eh1991 = ehCut.query("EH_SET == '1991'")
eh2001 = ehCut.query("EH_SET == '2001'")
mean1991 = round(eh1991[geo].values[0],3)
mean2001 = round(eh2001[geo].values[0],3)
sd1991 = eh1991[geo + '_SD'].values[0]
sd2001 = eh2001[geo + '_SD'].values[0]
title = geo + ' ' + aa + '\nEH 2001: mean=' + str(mean2001) + ' (' +str(sd2001) + ')\n'
title = title + 'EH 1991: mean=' + str(mean1991) + ' (' + str(sd1991) + ')'
print(aa,geo,mean1991,sd1991,mean2001,sd2001)
bestCut = dataBest.query("aa == '" + aa + "'")
if geo == 'N:CA:C':
georepAA.addHistogram(data=bestCut, geoX='TAU', title=title)
else:
georepAA.addHistogram(data=bestCut, geoX=geo, title=title)
'''
georepSummary.printToHtml(
'Best Supported Engh&Huber Compare, set=' + pdbSet, 4,
'Defensible_EH_' + pdbSet)
# -- ©Rachel Alcraft 2020, PsuGeometry --
from PsuGeometry import GeoReport as geor
'''
This script runs a correlation report on a few structures to demonstrate the use of 2Fo-Fc as a hue
The data is precalculated into a datafame
'''
pdbDataPath = '/home/rachel/Documents/Bioinformatics/ProteinDataFiles/pdb_data/'
edDataPath = '/home/rachel/Documents/Bioinformatics/ProteinDataFiles/ccp4_data/'
printPath = '/home/rachel/Documents/Bioinformatics/ProteinDataFiles/results_psu/Levels/'
pdbList = ['1ejg', '1us0', '1tt8', '1i1w', '1ucs', '6jvv', '5nqo']
georep = geor.GeoReport(pdbList,
pdbDataPath,
edDataPath,
printPath,
ed=True,
dssp=True)
geoList = [
'PHI', 'PSI', 'TAU', 'C-1:C', 'C-1:N:CA', 'CHI1', 'CA-1:CA', 'OMEGA',
'CA:CA+1', 'C-1:N:CA:C'
]
hueList = ['aa', 'bfactor', 'rid', 'resolution', 'pdbCode', '2FoFc', 'dssp']
data = georep.getGeoemtryCsv(geoList, hueList)
georep.addScatter(data=data,
geoX='PHI',
geoY='PSI',
hue='2FoFc',
