#! /usr/bin/env python2

__version__ = "1.5.0"

import os

import sys

import argparse

from itertools import combinations

from importlib import import_module

from distutils.sysconfig import get_python_lib

import matplotlib as mpl

mpl.use('agg')

from PyQt4 import QtGui, QtCore

from stamp.metagenomics.fileIO.StampIO import StampIO

from stamp.metagenomics.fileIO.MetadataIO import MetadataIO

from stamp.metagenomics.stats.GroupStatsTests import GroupStatsTests

from stamp.metagenomics.TableHelper import SortTableStrCol

from stamp.plugins.PlotsManager import PlotsManager

from stamp.plugins.groups.AbstractGroupPlotPlugin import AbstractGroupPlotPlugin

def load_data(profileFile, metadataFile):

return profileTree, metadata

# groupStatTestDict = loadPlugins(preferences, 'plugins/groups/statisticalTests/')

def loadPlugins(preferences, pluginFolder):

return d

# plotDict = loadPlots(preferences, 'plugins/groups/plots/')

def loadPlots(preferences, pluginFolder):

return plotDict

def initLegend(profileTree, metadata, field):

return groupColourDict

if __name__ == '__main__':

parser = argparse.ArgumentParser(formatter_class=lambda prog: argparse.ArgumentDefaultsHelpFormatter(prog, max_help_position=36, width=120))

parser.add_argument('-i', '--profile', required=True, help='Path to the profile file')

parser.add_argument('-g', '--metadata', required=True, help='Path to the group metadata file')

parser.add_argument('-d', '--output', default='output', help='Output directory')

parser.add_argument('-p', '--plot', default='{g1}-vs-{g2}.psig.png', help='Plot filename template, supported filetypes: png, pdf, ps, eps, svg')

parser.add_argument('-t', '--table', default='{g1}-vs-{g2}.test.xls', help='Table filename template')

parser.add_argument('-r', '--dpi', type=int, default=500, help='Plot DPI (dots per inch)')

parser.add_argument('-v', '--version', action='version', version='%(prog)s ' + __version__)

args = parser.parse_args()

# user input

profileFile = 'otu.g.spf'

metadataFile = 'Grouping.txt'

saveFolder = 'output'

savePlotTemplate = '{g1}-vs-{g2}.psig.png'

saveTableTemplate = '{g1}-vs-{g2}.test.xls'

dpi = 500

profileFile = args.profile

metadataFile = args.metadata

saveFolder = args.output

savePlotTemplate = args.plot

saveTableTemplate = args.table

dpi = args.dpi

# setup default plot settings

mpl.rcParams['font.size'] = 8

mpl.rcParams['axes.titlesize'] = 8

mpl.rcParams['axes.labelsize'] = 8

mpl.rcParams['xtick.labelsize'] = 8

mpl.rcParams['ytick.labelsize'] = 8

mpl.rcParams['legend.fontsize'] = 8

# initialize preferences

preferences = {}

settings = QtCore.QSettings("BeikoLab", "STAMP")

preferences['Settings'] = settings

preferences['Executable directory'] = sys.path[0]

preferences['Pseudocount'] = settings.value('Preferences/Pseudocount', 0.5).toDouble()[0]

preferences['Replicates'] = settings.value('Preferences/Replicates', 1000).toInt()[0]

preferences['Truncate feature names'] = settings.value('Preferences/Truncate feature names', True).toBool()

preferences['Length of truncated feature names'] = settings.value('Preferences/Length of truncated feature names', 50).toInt()[0]

preferences['Axes colour'] = QtGui.QColor(settings.value('Preferences/Axes colour', '#7f7f7f'))

preferences['All other samples colour'] = QtGui.QColor(settings.value('Preferences/All other samples colour', '#7f7f7f'))

preferences['Minimum reported p-value exponent'] = settings.value('Preferences/Minimum reported p-value exponent', -15).toDouble()[0]

preferences['Sample 1 colour'] = QtGui.QColor(128, 177, 211)

preferences['Sample 2 colour'] = QtGui.QColor(253, 180, 98)

preferences['Group colours'] = {}

preferences['Highlighted sample features'] = []

preferences['Highlighted group features'] = []

preferences['Highlighted multiple group features'] = []

preferences['Selected group feature'] = ''

preferences['Selected multiple group feature'] = ''

defaultPlot = 'PCA plot'

plotDict = loadPlots(preferences, 'plugins/groups/plots/')

groupPlotOptions = sorted(plotDict.keys())

# ['Bar plot', 'Box plot', 'Extended error bar', 'Heatmap plot', 'PCA plot', 'Scatter plot']

groupPlotIndex = 2

groupPlot = groupPlotOptions[groupPlotIndex]

profileTree, metadata = load_data(profileFile, metadataFile)

parentLevelOptions = ['Entire sample'] + profileTree.hierarchyHeadings

parentLevel = parentLevelOptions[0]

profileLevelOptions = profileTree.hierarchyHeadings

profileLevelIndex = 0

profileLevel = profileLevelOptions[profileLevelIndex]

unclassifiedTreatmentOptions = ['Retain unclassified reads',

'Remove unclassified reads',

'Use only for calculating frequency profiles']

unclassifiedTreatment = unclassifiedTreatmentOptions[0]

typeOfTestOptions = ['Multiple groups', 'Two groups', 'Two samples']

typeOfTest = typeOfTestOptions[1]

groupFieldOptions = metadata.getFeatures()

groupField = groupFieldOptions[0]

# setup group legend

groupColorDict = initLegend(profileTree, metadata, groupField)

preferences['Group colours'] = groupColorDict

samples = sorted(profileTree.sampleNames)

groups = sorted(profileTree.groupActive.keys())

# populateGroupComboBoxes

#groupName1 = groups[2]

#groupName2 = groups[4]

for groupName1, groupName2 in combinations(groups, 2):

#for groupName1, groupName2 in [(groups[0], groups[2])]:

# indicate the hierarchical level of interest has changed

## groupHierarchicalLevelsChanged()

groupHighlightHierarchyOptions = ['None'] + profileTree.hierarchyHeadings[0:profileLevelIndex + 1]

groupHighlightHierarchyIndex = 0

groupHighlightHierarchy = groupHighlightHierarchyOptions[groupHighlightHierarchyIndex]

groupHighlightFeatureOptions = []

groupEffectSizeDict = loadPlugins(preferences, 'plugins/groups/effectSizeFilters/')

# 'Ratio of proportions', 'Difference between proportions'

groupEffectSizeMeasure1Options = sorted(groupEffectSizeDict.keys())

groupEffectSizeMeasure1Index = 1

groupEffectSizeMeasure1 = groupEffectSizeMeasure1Options[groupEffectSizeMeasure1Index]

groupEffectSizeMeasure2Options = sorted(groupEffectSizeDict.keys())

groupEffectSizeMeasure2Index = 0

groupEffectSizeMeasure2 = groupEffectSizeMeasure2Options[groupEffectSizeMeasure2Index]

groupStatTestDict = loadPlugins(preferences, 'plugins/groups/statisticalTests/')

# "White's non-parametric t-test", "Welch's t-test", 't-test (equal variance)'

groupStatTest = "Welch's t-test"

multCompDict = loadPlugins(preferences, 'plugins/common/multipleComparisonCorrections/')

# 'Benjamini-Hochberg FDR', 'Sidak', 'Storey FDR', 'Bonferroni', 'No correction'

groupMultCompMethodOptions = sorted(multCompDict.keys())

# ['Benjamini-Hochberg FDR', 'Bonferroni', 'No correction', 'Sidak', 'Storey FDR']

groupMultCompMethodIndex = 2 # 'No correction'

groupMultCompMethod = groupMultCompMethodOptions[groupMultCompMethodIndex]

groupSignTestTypeOptions = ['One-sided', 'Two-sided']

groupSignTestTypeIndex = 1

groupSignTestType = groupSignTestTypeOptions[groupSignTestTypeIndex]

groupStatsTest = GroupStatsTests(preferences)

# run statistics

### groupRunTest()

# groupTestConfIntervMethods()

# populate combo box with CI methods compatible with current hypothesis test

test = groupStatTestDict[groupStatTest]

groupConfIntervMethodOptions = test.confIntervMethods

# ["DP: Welch's inverted"]

groupConfIntervMethod = groupConfIntervMethodOptions[0]

groupNominalCoverageOptions = ['0.90', '0.95', '0.98', '0.99', '0.999']

groupNominalCoverage = groupNominalCoverageOptions[1]

# show progress of test

print 'Running two-group statistical test ({0}, {1}),'.format(groupName1, groupName2),

# create profile

groupProfile = profileTree.createGroupProfile(groupName1, groupName2, parentLevel, profileLevel, metadata, unclassifiedTreatment)

# run significance test

groupStatsTest.run(test, groupSignTestType, groupConfIntervMethod, float(groupNominalCoverage), groupProfile)

# apply multiple test correction

groupStatsTest.results.performMultCompCorrection(multCompDict[groupMultCompMethod])

# apply filters

#### groupApplyFilters()

selectedFeatures = []

if not selectedFeatures:

groupStatsTest.results.selectAllFeautres()

selectedFeatures = groupStatsTest.results.getSelectedFeatures()

inactiveFeatures = ['Others']

selectedFeatures = list(set(selectedFeatures) - set(inactiveFeatures))

groupStatsTest.results.setSelectedFeatures(selectedFeatures)

# perform filtering

groupEnableSignLevelFilter = True

groupSignLevelFilter = 0.05

if not groupSignLevelFilter:

groupSignLevelFilter = None

# sequence filtering

groupEnableSeqFilter = False

groupSeqFilterOptions = ['maximum', 'minimum', 'independent, maximum', 'independent, minimum']

groupSeqFilterIndex = 0

groupSeqFilter = groupSeqFilterOptions[groupSeqFilterIndex]

group1Filter = 5

group2Filter = 5

if not groupEnableSeqFilter:

groupSeqFilter = None

group1Filter = None

group2Filter = None

groupEnableParentSeqFilter = False

groupParentSeqFilterOptions = ['maximum', 'minimum', 'independent, maximum', 'independent, minimum']

groupParentSeqFilterIndex = 0

groupParentSeqFilter = groupParentSeqFilterOptions[groupParentSeqFilterIndex]

groupParentGroup1Filter = 1

groupParentGroup2Filter = 1

if not groupEnableParentSeqFilter:

groupParentSeqFilter = None

groupParentGroup1Filter = None

groupParentGroup2Filter = None

# effect size filters

groupEnableEffectSizeFilter1 = False

if groupEnableEffectSizeFilter1:

groupMinEffectSize1 = 1.0

else:

groupEffectSizeMeasure1 = None

groupMinEffectSize1 = None

groupEnableEffectSizeFilter2 = False

if groupEnableEffectSizeFilter2:

groupMinEffectSize2 = 2.0

else:

groupEffectSizeMeasure2 = None

groupMinEffectSize2 = None

effectSizeOperatorOptions = ['OR', 'AND']

effectSizeOperatorIndex = 0

effectSizeOperator = effectSizeOperatorOptions[effectSizeOperatorIndex]

groupStatsTest.results.filterFeatures(groupSignLevelFilter, groupSeqFilter, group1Filter, group2Filter, groupParentSeqFilter, groupParentGroup1Filter, groupParentGroup2Filter, groupEffectSizeMeasure1, groupMinEffectSize1, effectSizeOperator, groupMinEffectSize2, groupMinEffectSize2)

# update table summarizing statistical results

#self.groupFeaturesTableUpdate()

#self.groupTable.updateTable(self.groupStatsTest)

# update plots

##### groupPlotUpdate()

# update plot

if groupStatsTest.results.activeData != []:

print 'Plotting:',

plotDict[groupPlot].sortingField = 'Effect sizes'

plotDict[groupPlot].figHeightPerRow = 0.55

plotDict[groupPlot].legendPos = 2

plotDict[groupPlot].bShowCorrectedPvalues = False

plotDict[groupPlot].plot(groupProfile, groupStatsTest.results)

# saveFolder = 'output'

# savePlotTemplate = '{g1}-vs-{g2}.png'

# saveTableTemplate = '{g1}-vs-{g2}.test.xls'

filename = os.path.join(saveFolder, savePlotTemplate.format(g1=groupName1, g2=groupName2))

print '"{0}",'.format(filename),

if not os.path.exists(os.path.dirname(filename)):

try:

os.makedirs(os.path.dirname(filename))

except OSError as exc: # Guard against race condition

if exc.errno != errno.EEXIST:

raise

ext = filename[filename.rfind('.')+1:len(filename)]

# dpi = 500

if ext in ['png', 'pdf', 'ps', 'eps','svg']:

plotDict[groupPlot].fig.savefig(filename, format=ext, dpi=dpi, facecolor='white', edgecolor='white')

else:

print 'Unsupported plot output format:', ext

# Prepare table

# ['Benjamini-Hochberg FDR', 'Bonferroni', 'No correction', 'Sidak', 'Storey FDR']

print 'Sidak table:',

groupMultCompMethodIndex = 3 # 'No correction'

groupMultCompMethod = groupMultCompMethodOptions[groupMultCompMethodIndex]

# run significance test

groupStatsTest.run(test, groupSignTestType, groupConfIntervMethod, float(groupNominalCoverage), groupProfile)

# apply multiple test correction

groupStatsTest.results.performMultCompCorrection(multCompDict[groupMultCompMethod])

groupStatsTest.results.setSelectedFeatures(selectedFeatures)

groupStatsTest.results.filterFeatures(groupSignLevelFilter, groupSeqFilter, group1Filter, group2Filter, groupParentSeqFilter, groupParentGroup1Filter, groupParentGroup2Filter, groupEffectSizeMeasure1, groupMinEffectSize1, effectSizeOperator, groupMinEffectSize2, groupMinEffectSize2)

tableData, tableHeadings = groupStatsTest.results.tableData(False)

tableData = SortTableStrCol(tableData, 0)

filename = os.path.join(saveFolder, saveTableTemplate.format(g1=groupName1, g2=groupName2))

print '"{0}"'.format(filename),

fout = open(filename, 'w')

for heading in tableHeadings:

fout.write(heading + '\t')

fout.write('\n')

for row in tableData:

for entry in row:

fout.write(str(entry) + '\t')

fout.write('\n')

fout.close()

print 'DONE.'

else:

print 'No Active Data, SKIP.'