def checkPathBarriers(prefix):
tstName = prefix + ".nocav.tst"
findHolesName = tstName + ".findholes.log"
findHolesFile = open(findHolesName, 'r')
findHolesLines = findHolesFile.readlines()
findHolesFile.close()
HolesName = tstName + ".sideshole.log" # holds all the output
goodHolesName = tstName + ".good.sideshole.log" # just the 1 1 0 1 1
sideHolesName = tstName + ".side.sideshole.log" # just the * * 1 * *
badHolesName = tstName + ".bad.sideshole.log" # all others
pdbWithBarriersFileName = "planes_" + prefix + ".pdb"
pdbBarriers = pdb.pdbData(pdbWithBarriersFileName)
#get the barriers read in and defined
barrierAtomList = [[], []]
for index, resName in enumerate(pdbBarriers.resNames):
if resName == "DUM":
if pdbBarriers.atoms[index][0] == "O":
barrierAtomList[0].append(pdbBarriers.coords[index])
elif pdbBarriers.atoms[index][0] == "N":
barrierAtomList[1].append(pdbBarriers.coords[index])
barrierZ = [barrierAtomList[0][0][2], barrierAtomList[1][0][2]]
barrierZ.sort()
barrierSep = geometry.distL2(barrierAtomList[0][0], barrierAtomList[1][0])
#barrier is just Z coordinate
#setup for main loop over paths
poreSuffix = ".pore.py"
logFile = open(HolesName, 'w')
goodLogFile = open(goodHolesName, 'w')
sideLogFile = open(sideHolesName, 'w')
badLogFile = open(badHolesName, 'w')
#the following 5 things are calculated and written for each path, headers
#the 6th, barrier separation, is really the same for each structure
logFile.write("endsBeyond1count barrier1count endsBetweenCount ")
logFile.write("barrier2count endsBeyond2count barrierSeparation\n")
goodLogFile.write("prefix ")
goodLogFile.write(string.strip(findHolesLines[0]) + " ")
goodLogFile.write("endsBeyond1count barrier1count endsBetweenCount ")
goodLogFile.write("barrier2count endsBeyond2count barrierSeparation\n")
sideLogFile.write("prefix ")
sideLogFile.write(string.strip(findHolesLines[0]) + " ")
sideLogFile.write("endsBeyond1count barrier1count endsBetweenCount ")
sideLogFile.write("barrier2count endsBeyond2count barrierSeparation\n")
badLogFile.write("prefix ")
badLogFile.write(string.strip(findHolesLines[0]) + " ")
badLogFile.write("endsBeyond1count barrier1count endsBetweenCount ")
badLogFile.write("barrier2count endsBeyond2count barrierSeparation\n")
holeNumber = 1
poreFile = tstName + "." + str(holeNumber) + poreSuffix
print poreFile
paths = []
sides, goods = [], []
endsToPaths = {}
pathsToEnds = {}
while os.path.exists(poreFile):
path = comparePaths.readCGOPath(poreFile)
pathRad = comparePaths.readCGOPathWithRadius(poreFile)
paths.append(pathRad)
pathNum = len(paths) - 1
for end in string.split(findHolesLines[holeNumber])[1:3]:
if pathNum not in pathsToEnds:
pathsToEnds[pathNum] = []
pathsToEnds[pathNum].append(end)
if end not in endsToPaths:
endsToPaths[end] = []
endsToPaths[end].append(pathNum)
intersections = [0, 0]
for index, barrier in enumerate(barrierZ):
intersections[index] = countCrossingsZ(path, barrier)
ends = [0, 0, 0]
for endPoint in [path[0], path[-1]]:
endPointZ = endPoint[2]
if endPointZ < barrierZ[0] and endPointZ < barrierZ[1]:
ends[0] += 1
elif endPointZ >= barrierZ[0] and endPointZ <= barrierZ[1]:
ends[1] += 1
elif endPointZ > barrierZ[0] and endPointZ > barrierZ[1]:
ends[2] += 1
outputThisTime = str(ends[0]) + " " + str(intersections[0]) + " " + \
str(ends[1]) + " " + str(intersections[1]) + " " + \
str(ends[2]) + " " + str(barrierSep) + " "
logFile.write(outputThisTime)
logFile.write("\n")
if ends[0] + ends[1] + ends[2] != 2:
print "problems sorting out the ends"
if ends[0] == 1 and ends[2] == 1 and intersections == [1, 1]:
#it is 'good'
goods.append(pathNum)
goodLogFile.write(prefix + " ")
goodLogFile.write(string.strip(findHolesLines[holeNumber]) + " ")
goodLogFile.write(outputThisTime + "\n")
elif ends[1] == 1:
sides.append(pathNum)
sideLogFile.write(prefix + " ")
sideLogFile.write(string.strip(findHolesLines[holeNumber]) + " ")
sideLogFile.write(outputThisTime + "\n")
else:
badLogFile.write(prefix + " ")
badLogFile.write(string.strip(findHolesLines[holeNumber]) + " ")
badLogFile.write(outputThisTime + "\n")
#and that is it for this path
holeNumber += 1 # get set up for next pass
poreFile = tstName + "." + str(holeNumber) + poreSuffix
logFile.close()
goodLogFile.close()
sideLogFile.close()
badLogFile.close()
#next lines are for debugging the new data structures
'''
print sides
print goods
print endsToPaths
print pathsToEnds
'''
#now want to find side branches of good paths
branches = 0
branchSuffix = ".branch.py"
branchFile = tstName + "." + str(branches) + branchSuffix
branchLog = open(tstName + ".branchholes.log", 'w')
branchLog.write(string.strip(findHolesLines[0]) + "\n")
for side in sides:
foundGoods = []
for sideEnd in pathsToEnds[side]:
for good in goods:
for goodEnd in pathsToEnds[good]:
if goodEnd == sideEnd:
foundGoods.append(good)
if len(foundGoods) > 0:
branchedPath = paths[side] # start with whole path
for good in foundGoods: # remove physiological intersecting paths
branchedPath = pathsModule.subtractPaths(branchedPath, paths[good])
if len(branchedPath) > 0: # has to have some length remaining
branches += 1
branchFile = tstName + "." + str(branches) + branchSuffix
print branches, side, foundGoods
tstdebug.debugSetGridSpheres(
branchedPath, 0.5, branchFile, radius=True,
mainColor=(0.01, 0.9, 0.35))
branchLog.write(str(branches) + " ")
branchLog.write(str(pathsToEnds[side][0]) + " ")
branchLog.write(str(pathsToEnds[side][1]) + " ")
branchLog.write("- ") # dummy, not real
branchLog.write("0. 0. 0. 0. 0. 0. 0. 0. 0. 0. \n")
branchLog.close()
addFoundHoleStats.redoFindholes(
prefix, nearbyDistance=4., logExt=".branchholes.log",
poreSuffix=".branch.py", nearbyName=".branch")
def tstTravelFindHoles(
tstFileName, phiFileName, debugOut=False, borderSize=2, nearbyDistance=4.):
'''if debugout is set, additional files are created.
bordersize can change the amount of extra space around the protein.
nearbydistance changes the distance that nearby residues are gathered from.'''
print "reading tst and phi files"
tstD = tstdata.tstData(
tstFileName, necessaryKeys=tstdata.tstData.necessaryKeysForHoles)
phiData = phi(phiFileName) # read in the phimap if possible
gridSize = 1.0 / phiData.scale # needed later
numberHandles = tstD.countHandles()
print "there are ", numberHandles, " holes in this structure"
print "running travel depth now"
phiTravelDepthGrid, phiTravelDepthData, meshData = tstTravelDepthMeshRun(
tstD, phiData, tstFileName, borderSize=borderSize, threshold="auto")
del phiData, phiTravelDepthGrid, phiTravelDepthData
#not needed, reclaim memory
print "calculating travel out distance"
meshData.calculateTravelDistance("surfout", [3], [0, 2])
print "finding holes"
loopTrisSave, loopPointsSave, regLoopTris, regLoopPts, pointNeighbors, \
pointNeighborsNodes, outsidePoints, outsidePointsNodes, possHoleStarts = \
tstTopology.fillInHolesAndGrow(
tstD.dict['TRIANGLE_POINT'], tstD.dict['POINT_TRIANGLE'],
tstD.dict['POINT_XYZ'], tstD.dict['NORM_XYZ'], numberHandles,
tstFileName, debugOut, meshData, "surfout")
if debugOut:
tstdebug.debugTriangleList(
regLoopTris, tstD.dict['TRIANGLE_POINT'], tstD.dict['POINT_XYZ'],
tstFileName + ".regular.loops.py")
tstDPointXYZ = tstD.dict['POINT_XYZ'] # save for later
tstDPdbRecord = tstD.dict['PDB_RECORD'] # save for later
del tstD # rest of tstD isn't needed, so free memory
#output the possible places where HOLE could start... centers of regular plugs
print "writing output files"
paths.outputNodesText(possHoleStarts, tstFileName + ".HOLE.start.txt")
pathsList = meshData.getPaths(
"surfout", pointNeighbors, outsidePoints, possHoleStarts)
del meshData # deletes everything that has no other refs (from paths)
#print len(pointNeighbors), len(outsidePoints), len(possHoleStarts),
#print len(pathsList)
allPoints, outsidePts = [], []
for point in pointNeighborsNodes.keys():
allPoints.append(point.pathXyz)
for point in outsidePointsNodes:
outsidePts.append(point.pathXyz)
tstdebug.pointDebug(allPoints, filename=tstFileName+".tree.py")
tstdebug.pointDebug(
outsidePts, filename=tstFileName+".outside.py", mainColor=(.9, .1, .1),
radius=0.55)
#tstdebug.debugSetGridSpheres(
# pointNeighbors.keys(),, gridSize, tstFileName+".tree.radius.py",
# radius=True, mainColor=(0.01, 0.9, 0.05))
#testing new output of tree with radius
foundPaths = 0
pathFile = string.replace(tstFileName, ".nocav.tst", ".py")
#very very specific... probably bad but nice...
#can always rerun standalone later.
knownPathExists = os.path.isfile(pathFile)
logName = tstFileName + ".findholes.log"
logFile = open(logName, 'w')
logFile.write(
"number endNumOne endNumTwo plugs stepLength pathLength pathMinRadius " +
"pathMaxInsideRadius endMinOne endMinTwo minimaCount travelDepthMax " +
"windingMetric avgTheta ")
if knownPathExists:
logFile.write("pRMSD coverage span wrmsd less1 lessrad radiicomp")
logFile.write("\n")
if knownPathExists:
bestStats = ["err", "err", "err", "err", "err", "err", "err"]
bestStatsPaths = [0, 0, 0, 0, 0, 0, 0]
sortedByMinRadiusPaths = []
for pathIndex, (outsideOne, outsideTwo, plugs, nodePath) in enumerate(
pathsList):
pointPath, spheres = [], []
for node in nodePath:
pointPath.append(list(node.pathXyz))
spheres.append(list(node.pathXyz))
pointRadius = node.distances["surfout"] # add 'radius' info
if not pointRadius or pointRadius == 0.:
pointRadius = .000000001 # very small number, on surface
pointPath[-1].insert(0, pointRadius)
spheres[-1].append(pointRadius)
minRad = paths.pathMinRadius(pointPath)
newTuple = (
minRad, outsideOne, outsideTwo, plugs, nodePath, pointPath, spheres)
#insertion sort into new list
position = 0
while position < len(sortedByMinRadiusPaths) and \
sortedByMinRadiusPaths[position][0] > minRad:
#only if list can handle it and already inserted are bigger
position += 1
sortedByMinRadiusPaths.insert(position, newTuple)
print "output files for individual paths"
for pathIndex, newTuple in enumerate(sortedByMinRadiusPaths):
(minRad, outsideOne, outsideTwo, plugs, nodePath, pointPath, spheres) = \
newTuple # unpack the tuple into the various data
throughTris, throughPts = paths.checkPath(
pointPath, loopPointsSave, tstDPointXYZ)
if throughTris: # it worked... make some more debugging output
foundPaths += 1
outName = tstFileName + "." + str(foundPaths)
if debugOut and throughTris:
tstdebug.debugTrianglesNotOrig(
throughTris, tstDPointXYZ, outName+".through.loop.py",
ptList=throughPts)
#always do these 2
tstdebug.debugSetGridSpheres(
pointPath, gridSize, outName + ".pore.py", radius=True,
mainColor=(0.01, 0.9, 0.05))
tstdebug.debugSetGridSpheres(
pointPath, gridSize, outName + ".path.py", mainColor=(.01, 0.95, 0.9))
#mesh.meshFromSpheres(spheres, 0.5, outName + ".points.py")
paths.outputRadiiTxt(pointPath, outName + ".radii.txt")
paths.outputNearbyResidues(
pointPath, outName, tstDPdbRecord, nearbyDistance)
pathLen = paths.pathLength(pointPath)
minimaCount = paths.pathMinimaCount(pointPath)
maxRad, endMinOne, endMinTwo = paths.insideTwoMinimaRadiusMax(pointPath)
travelDepthMax = paths.pathMaxDistance(nodePath, 'traveldepth')
windingMetric = paths.computeWindingMetric(pointPath)
thetas, avgTheta = paths.averageTheta(pointPath)
#print endMinOne, endMinTwo, minimaCount, travelDepthMax, windingMetric,
#print avgTheta
#attempt to do pRMSD if possible...
prmsd, coverage, span, wrmsd, less1, lessrad, radiicomp = \
"err", "err", "err", "err", "err", "err", "err"
if knownPathExists:
try:
sourcePath, sourceRadii = [], []
for pathPt in pointPath:
sourcePath.append(pathPt[1:4])
sourceRadii.append(pathPt[0])
prmsd, coverage, span, wrmsd, less1, lessrad, radiicomp = \
comparePathsManyMetrics(False, pathFile, sourcePath, sourceRadii)
if bestStats[0] == 'err' or bestStats[0] > prmsd:
bestStats[0] = prmsd
bestStatsPaths[0] = foundPaths
if bestStats[1] == 'err' or bestStats[1] < coverage:
bestStats[1] = coverage
bestStatsPaths[1] = foundPaths
if bestStats[2] == 'err' or bestStats[2] < span:
bestStats[2] = span
bestStatsPaths[2] = foundPaths
if bestStats[3] == 'err' or bestStats[3] > wrmsd:
bestStats[3] = wrmsd
bestStatsPaths[3] = foundPaths
if bestStats[4] == 'err' or bestStats[4] < less1:
bestStats[4] = less1
bestStatsPaths[4] = foundPaths
if bestStats[5] == 'err' or bestStats[5] < lessrad:
bestStats[5] = lessrad
bestStatsPaths[5] = foundPaths
if bestStats[6] == 'err' or bestStats[6] < radiicomp:
bestStats[6] = radiicomp
bestStatsPaths[6] = foundPaths
except (IOError, TypeError):
#if there is no known path file, this should be the error
pass
#now output data
logFile.write(str(foundPaths) + " ")
logFile.write(str(outsideOne) + " ")
logFile.write(str(outsideTwo) + " ")
logFile.write(str(plugs) + " ")
logFile.write(str(len(pointPath)) + " ")
logFile.write(str(pathLen) + " ")
logFile.write(str(minRad) + " ")
logFile.write(str(maxRad) + " ")
logFile.write(str(endMinOne) + " ")
logFile.write(str(endMinTwo) + " ")
logFile.write(str(minimaCount) + " ")
logFile.write(str(travelDepthMax) + " ")
logFile.write(str(windingMetric) + " ")
logFile.write(str(avgTheta) + " ")
if knownPathExists:
logFile.write(str(prmsd) + " ")
logFile.write(str(coverage) + " ")
logFile.write(str(span) + " ")
logFile.write(str(wrmsd) + " ")
logFile.write(str(less1) + " ")
logFile.write(str(lessrad) + " ")
logFile.write(str(radiicomp) + " ")
logFile.write("\n") # that's all
logFile.close()
if knownPathExists: # output bestStats and bestStatsPaths
bestName = tstFileName + ".known.best.txt"
bestFile = open(bestName, 'w')
bestFile.write("pRMSD coverage span wrmsd less1 lessrad radiicomp ")
bestFile.write("pRMSD# coverage# span# wrmsd# less1# lessrad# radiicomp#\n")
for stat in bestStats:
bestFile.write(str(stat) + " ")
for stat in bestStatsPaths:
bestFile.write(str(stat) + " ")
bestFile.write("\n")
bestFile.close()
print "done with chunnel"