sifFile = args[0]
featureFile = args[1]
iplMatrix = None
global verbose, globalList, globalPathway
for o, a in opts:
if o == "-q":
verbose = False
elif o == "-i":
iplMatrix = a
## build sourceList
typeFile = "/".join(re.split("/", sifFile)[:-2] + ["TYPE.NA"])
sourceList = re.split("/", sifFile)[-2] + ".list_t"
h = mData.rList(featureFile)
(n, i) = mPathway.rSIF(sifFile, typef=typeFile)
(gn, gi) = mPathway.rPathway(globalPathway)
p = mPathway.Pathway(n, i)
s = mPathway.sortConnected(p)
f = open(sourceList, "w")
c = 1
for i in s:
u = list(set(i) & set(h))
if len(u) >= 5:
f.write("component_%s\t%s\n" % (c, "\t".join(u)))
c += 1
break
f.write("component_all\t%s\n" % ("\t".join(list(set(n.keys()) & set(h)))))
f.write("all\t%s\n" % ("\t".join(list(set(gn.keys()) & set(h)))))
f.close()
def cli_routine(outputDir, circleFiles, orderFiles, sampleFile, featureFile, orderFeature, centerFile, colorscaleFile, printLabel, verbose, cohortMinMax=False, purpleHack = True): """Routine for program execution via command-line.""" # I've tried not to touch this method as much as possible. # I don't want to break the way it was working for Sam Ng. # chrisw ## execute samples = [] features = [] if sampleFile != None: samples = mData.rList(sampleFile) if featureFile != None: features = mData.rList(featureFile) # end section for getting lists of samples and features ## read circleFiles # circleData is a list of dict[col][row]=score from each circleFile circleData = [] # circleColorsPalette is a list of (minColor),(zeroColor),(maxColor) circleColorsPalette = [] ## read colorscaleFile # the format is as follows - header compulsory: # min/max color coding color1 color2 color 3 # -2,2 rgb 155,155,155 255,255,255 0,0,0, # - rgb 155,0,155 255,0,255 0,0,0, # the "color format" is intended to support more color format, as I have # seen the html-colors in the code. # Michael ([email protected]) colorscaleData = None if colorscaleFile != None: if cohortMinMax: log("WARNING: The -k option overrides -m") colorscaleData = mData.retRows(colorscaleFile,aslist=True) line=1 for cs in colorscaleData: line = line + 1 if len(cs) != 5: log("ERROR: color scale needs five fields: datapoints, colorcoding(rgb) and three colors\n", die = True) try: cs[0] = [float(x) for x in cs[0].split(",")] except ValueError: pass if len(cs[0]) != 2 and cs[0] != "-": print cs[0] log("ERROR: Two data points or dash needed for color scale\n", die = True) if cs[1].lower() == "rgb": try: cs[2] = rgb(*[float(x) for x in cs[2].split(",")]) cs[3] = rgb(*[float(x) for x in cs[3].split(",")]) cs[4] = rgb(*[float(x) for x in cs[4].split(",")]) except TypeError: log("ERROR: RGB needs three values on line " + str(line) + "\n", die = True) except ValueError: log("ERROR: RGB color not correctly defined on line " + str(line) + "\n", die=True) else: log("ERROR: Unknown color coding on line " + str(line) + ": " + str(cs[1]) + "\n", die=True) for i in xrange(len(circleFiles)): # get data, samples, and features from each circleFile # data is a dict[col][row]=score # cols is a list of sample names # features is a list of feature names (data, cols, rows) = mData.rCRSData(circleFiles[i], retFeatures=True) circleData.append(data) minCol = lightBlueRGB zerCol = whiteRGB maxCol = redRGB if colorscaleFile != None and i<len(colorscaleData): #get colors from specified colorscaleFile minCol = colorscaleData[i][2] zerCol = colorscaleData[i][3] maxCol = colorscaleData[i][4] # special cases for -meth and -mut # if circleFiles[i].endswith("meth"): # maxCol = blueRGB # minCol = redRGB # log("Color: meth\n") # elif circleFiles[i].endswith("mut"): # maxCol = blackRGB # minCol = whiteRGB # log("Color: mut\n") circleColorsPalette.append((minCol, zerCol, maxCol)) # if no sampleFile/featureFile, default to using samples/features from circleFiles if sampleFile == None: samples = list(set(cols) | set(samples)) if featureFile == None: features = list(set(rows) | set(features)) # end section for reading circleFiles ## read centerFile centerData = None if centerFile != None: centerData = mData.r2Col(centerFile, header=True) ## sort if orderFeature != None: if len(orderFiles) > 0: orderData = [] orderColors = [] for i in xrange(len(orderFiles)): orderData.append(mData.rCRSData(orderFiles[i])) minCol = whiteRGB zerCol = whiteRGB maxCol = blackRGB orderColors.append((minCol, zerCol, maxCol)) else: orderData = circleData # sort samples based on sample score in orderData # priority of sorting determined by orderFiles parameter samples.sort(lambda x, y: scmp(x, y, orderFeature, orderData)) ## cohort png # cgi will probably not use orderFiles if len(orderFiles) > 0: imgFile = "%s/Cohort.png" % (outputDir) label = "Cohort" centerCol = whiteRGB.tohex() cohortCircleCols = [] for i in xrange(len(orderData)): ringCols = [] ringVals = [] for sample in samples: if sample in orderData[i]: if orderFeature in orderData[i][sample]: ringVals.append(orderData[i][sample][orderFeature]) elif "*" in orderData[i][sample]: ringVals.append(orderData[i][sample]["*"]) minVal = min([-0.01] + mData.floatList(ringVals)) maxVal = max([0.01] + mData.floatList(ringVals)) for sample in samples: if sample in orderData[i]: if orderFeature in orderData[i][sample]: ringCols.append(getColor(orderData[i][sample][orderFeature], minVal, maxVal, minColor=orderColors[i][0], zeroColor=orderColors[i][1], maxColor=orderColors[i][2])) elif "*" in orderData[i][sample]: ringCols.append(getColor(orderData[i][sample]["*"], minVal, maxVal, minColor=orderColors[i][0], zeroColor=orderColors[i][1], maxColor=orderColors[i][2])) else: ringCols.append(greyRGB.tohex()) else: ringCols.append(greyRGB.tohex()) cohortCircleCols.append(ringCols) plotCircle(imgFile, label=label, centerCol=centerCol, circleCols=cohortCircleCols, innerRadTotal=0.2, outerRadTotal=0.5, width=5) # end section for sample ordering ## plot images if centerData != None: centerDataFloatList = mData.floatList(centerData.values()) centerDataMinVal = min([-0.01] + centerDataFloatList) centerDataMaxVal = max([0.01] + centerDataFloatList) # get min/max values for datasets if cohortMinMax: (minValList, maxValList) = getCohortMinMaxValues(features, samples, circleData) else: (minValList, maxValList) = (None, None) if colorscaleData != None: (minValList, maxValList) = getColorScaleMinMaxValues(minValList, maxValList, len(circleData), colorscaleData) for feature in features: log("Drawing %s\n" % (feature)) centerColHex = None if centerData != None: if feature in centerData: centerColHex = getColor(centerData[feature], centerDataMinVal, centerDataMaxVal, minColor=lightBlueRGB, zeroColor=whiteRGB, purple0Hack=purpleHack) imgFile = "%s/%s.png" % (outputDir, re.sub("[/:]", "_", feature)) label = "" if printLabel: label = feature image_width = 5.0 drawCircleImageForFeature(feature, samples, label, imgFile, circleData, circleColorsPalette, width=image_width, centerColHex=centerColHex, minValList=minValList, maxValList=maxValList, purple0Hack=purpleHack) for sample in samples: log("ordered samples: %s\n" % (sample))
orderFiles = []
else:
orderFeature = sa[0]
orderFiles = re.split(",", sa[1])
elif o == "-c":
centerFile = a
elif o == "-l":
printLabel = True
elif o == "-q":
verbose = False
## execute
samples = []
features = []
if sampleFile != None:
samples = mData.rList(sampleFile)
if featureFile != None:
features = mData.rList(featureFile)
## read circleFiles
circleData = []
circleColors = []
for i in range(len(circleFiles)):
(data, cols, rows) = mData.rCRSData(circleFiles[i], retFeatures = True)
circleData.append(data)
minCol = rgb(0, 0, 255)
zerCol = rgb(255, 255, 255)
maxCol = rgb(255, 0, 0)
if circleFiles[i].endswith("meth"):
maxCol = rgb(0, 0, 255)
minCol = rgb(255, 0, 0)
sifFile = args[0]
featureFile = args[1]
iplMatrix = None
global verbose, globalList, globalPathway
for o, a in opts:
if o == "-q":
verbose = False
elif o == "-i":
iplMatrix = a
## build sourceList
typeFile = "/".join(re.split("/", sifFile)[:-2] + ["TYPE.NA"])
sourceList = re.split("/", sifFile)[-2] + ".list_t"
h = mData.rList(featureFile)
(n, i) = mPathway.rSIF(sifFile, typef=typeFile)
(gn, gi) = mPathway.rPathway(globalPathway)
p = mPathway.Pathway(n, i)
s = mPathway.sortConnected(p)
f = open(sourceList, "w")
c = 1
for i in s:
u = list(set(i) & set(h))
if len(u) >= 5:
f.write("component_%s\t%s\n" % (c, "\t".join(u)))
c += 1
break
f.write("component_all\t%s\n" % ("\t".join(list(set(n.keys()) & set(h)))))
f.write("all\t%s\n" % ("\t".join(list(set(gn.keys()) & set(h)))))
f.close()
import os, re
import mData
assert (os.path.exists("first.props"))
assert (os.path.exists("include.features"))
features = mData.rList("include.features")
f = open("first.props", "r")
o = open("all.props", "w")
for line in f:
if line.startswith(
"nodeCustomGraphics1.default-Node\ Custom\ Graphics\ 1-Discrete\ Mapper.mapping.map"
):
pline = re.split(",", line.rstrip("\r\n"))
imgCounter = int(pline[1])
for feature in features:
reline = re.sub(features[0], feature,
pline[0]) + "," + str(imgCounter) + "," + re.sub(
features[0], feature,
pline[1]) + "," + pline[2]
imgCounter += 1
else:
o.write(line)
f.close()
o.close()
def cli_routine(outputDir,
circleFiles,
orderFiles,
sampleFile,
featureFile,
orderFeature,
centerFile,
colorscaleFile,
printLabel,
verbose,
cohortMinMax=False,
purpleHack=True):
"""Routine for program execution via command-line."""
# I've tried not to touch this method as much as possible.
# I don't want to break the way it was working for Sam Ng.
# chrisw
## execute
samples = []
features = []
if sampleFile != None:
samples = mData.rList(sampleFile)
if featureFile != None:
features = mData.rList(featureFile)
# end section for getting lists of samples and features
## read circleFiles
# circleData is a list of dict[col][row]=score from each circleFile
circleData = []
# circleColorsPalette is a list of (minColor),(zeroColor),(maxColor)
circleColorsPalette = []
## read colorscaleFile
# the format is as follows - header compulsory:
# min/max color coding color1 color2 color 3
# -2,2 rgb 155,155,155 255,255,255 0,0,0,
# - rgb 155,0,155 255,0,255 0,0,0,
# the "color format" is intended to support more color format, as I have
# seen the html-colors in the code.
# Michael ([email protected])
colorscaleData = None
if colorscaleFile != None:
if cohortMinMax:
log("WARNING: The -k option overrides -m")
colorscaleData = mData.retRows(colorscaleFile, aslist=True)
line = 1
for cs in colorscaleData:
line = line + 1
if len(cs) != 5:
log("ERROR: color scale needs five fields: datapoints, colorcoding(rgb) and three colors\n",
die=True)
try:
cs[0] = [float(x) for x in cs[0].split(",")]
except ValueError:
pass
if len(cs[0]) != 2 and cs[0] != "-":
print cs[0]
log("ERROR: Two data points or dash needed for color scale\n",
die=True)
if cs[1].lower() == "rgb":
try:
cs[2] = rgb(*[float(x) for x in cs[2].split(",")])
cs[3] = rgb(*[float(x) for x in cs[3].split(",")])
cs[4] = rgb(*[float(x) for x in cs[4].split(",")])
except TypeError:
log("ERROR: RGB needs three values on line " + str(line) +
"\n",
die=True)
except ValueError:
log("ERROR: RGB color not correctly defined on line " +
str(line) + "\n",
die=True)
else:
log("ERROR: Unknown color coding on line " + str(line) + ": " +
str(cs[1]) + "\n",
die=True)
for i in xrange(len(circleFiles)):
# get data, samples, and features from each circleFile
# data is a dict[col][row]=score
# cols is a list of sample names
# features is a list of feature names
(data, cols, rows) = mData.rCRSData(circleFiles[i], retFeatures=True)
circleData.append(data)
minCol = lightBlueRGB
zerCol = whiteRGB
maxCol = redRGB
if colorscaleFile != None and i < len(colorscaleData):
#get colors from specified colorscaleFile
minCol = colorscaleData[i][2]
zerCol = colorscaleData[i][3]
maxCol = colorscaleData[i][4]
# special cases for -meth and -mut
# if circleFiles[i].endswith("meth"):
# maxCol = blueRGB
# minCol = redRGB
# log("Color: meth\n")
# elif circleFiles[i].endswith("mut"):
# maxCol = blackRGB
# minCol = whiteRGB
# log("Color: mut\n")
circleColorsPalette.append((minCol, zerCol, maxCol))
# if no sampleFile/featureFile, default to using samples/features from circleFiles
if sampleFile == None:
samples = list(set(cols) | set(samples))
if featureFile == None:
features = list(set(rows) | set(features))
# end section for reading circleFiles
## read centerFile
centerData = None
if centerFile != None:
centerData = mData.r2Col(centerFile, header=True)
## sort
if orderFeature != None:
if len(orderFiles) > 0:
orderData = []
orderColors = []
for i in xrange(len(orderFiles)):
orderData.append(mData.rCRSData(orderFiles[i]))
minCol = whiteRGB
zerCol = whiteRGB
maxCol = blackRGB
orderColors.append((minCol, zerCol, maxCol))
else:
orderData = circleData
# sort samples based on sample score in orderData
# priority of sorting determined by orderFiles parameter
samples.sort(lambda x, y: scmp(x, y, orderFeature, orderData))
## cohort png
# cgi will probably not use orderFiles
if len(orderFiles) > 0:
imgFile = "%s/Cohort.png" % (outputDir)
label = "Cohort"
centerCol = whiteRGB.tohex()
cohortCircleCols = []
for i in xrange(len(orderData)):
ringCols = []
ringVals = []
for sample in samples:
if sample in orderData[i]:
if orderFeature in orderData[i][sample]:
ringVals.append(orderData[i][sample][orderFeature])
elif "*" in orderData[i][sample]:
ringVals.append(orderData[i][sample]["*"])
minVal = min([-0.01] + mData.floatList(ringVals))
maxVal = max([0.01] + mData.floatList(ringVals))
for sample in samples:
if sample in orderData[i]:
if orderFeature in orderData[i][sample]:
ringCols.append(
getColor(orderData[i][sample][orderFeature],
minVal,
maxVal,
minColor=orderColors[i][0],
zeroColor=orderColors[i][1],
maxColor=orderColors[i][2]))
elif "*" in orderData[i][sample]:
ringCols.append(
getColor(orderData[i][sample]["*"],
minVal,
maxVal,
minColor=orderColors[i][0],
zeroColor=orderColors[i][1],
maxColor=orderColors[i][2]))
else:
ringCols.append(greyRGB.tohex())
else:
ringCols.append(greyRGB.tohex())
cohortCircleCols.append(ringCols)
plotCircle(imgFile,
label=label,
centerCol=centerCol,
circleCols=cohortCircleCols,
innerRadTotal=0.2,
outerRadTotal=0.5,
width=5)
# end section for sample ordering
## plot images
if centerData != None:
centerDataFloatList = mData.floatList(centerData.values())
centerDataMinVal = min([-0.01] + centerDataFloatList)
centerDataMaxVal = max([0.01] + centerDataFloatList)
# get min/max values for datasets
if cohortMinMax:
(minValList,
maxValList) = getCohortMinMaxValues(features, samples, circleData)
else:
(minValList, maxValList) = (None, None)
if colorscaleData != None:
(minValList,
maxValList) = getColorScaleMinMaxValues(minValList, maxValList,
len(circleData),
colorscaleData)
for feature in features:
log("Drawing %s\n" % (feature))
centerColHex = None
if centerData != None:
if feature in centerData:
centerColHex = getColor(centerData[feature],
centerDataMinVal,
centerDataMaxVal,
minColor=lightBlueRGB,
zeroColor=whiteRGB,
purple0Hack=purpleHack)
imgFile = "%s/%s.png" % (outputDir, re.sub("[/:]", "_", feature))
label = ""
if printLabel:
label = feature
image_width = 5.0
drawCircleImageForFeature(feature,
samples,
label,
imgFile,
circleData,
circleColorsPalette,
width=image_width,
centerColHex=centerColHex,
minValList=minValList,
maxValList=maxValList,
purple0Hack=purpleHack)
for sample in samples:
log("ordered samples: %s\n" % (sample))
import os, re
import mData
assert(os.path.exists("first.props"))
assert(os.path.exists("include.features"))
features = mData.rList("include.features")
f = open("first.props", "r")
o = open("all.props", "w")
for line in f:
if line.startswith("nodeCustomGraphics1.default-Node\ Custom\ Graphics\ 1-Discrete\ Mapper.mapping.map"):
pline = re.split(",", line.rstrip("\r\n"))
imgCounter = int(pline[1])
for feature in features:
reline = re.sub(features[0], feature, pline[0]) + "," + str(imgCounter) + "," + re.sub(features[0], feature, pline[1]) + "," + pline[2]
imgCounter += 1
else:
o.write(line)
f.close()
o.close()
orderFiles = []
else:
orderFeature = sa[0]
orderFiles = re.split(",", sa[1])
elif o == "-c":
centerFile = a
elif o == "-l":
printLabel = True
elif o == "-q":
verbose = False
## execute
samples = []
features = []
if sampleFile != None:
samples = mData.rList(sampleFile)
if featureFile != None:
features = mData.rList(featureFile)
## read circleFiles
circleData = []
circleColors = []
##
## record file types for each, effects the color scheme
## use the input index for each
color_scheme_map = {}
for i in range(len(circleFiles)):
circleFile, colorScheme = circleFiles[i].split(':')
color_scheme_map[i] = parseColorScheme(colorScheme)
(data, cols, rows) = mData.rCRSData(circleFile, retFeatures = True)
circleData.append(data)
