#!/usr/bin/env python
# -*- coding: utf-8 -*-
#
#       KEGGFisher.py
#       
#       Copyright 2013 Oscar Reina <oscar@niska.bx.psu.edu>
#       
#       This program is free software; you can redistribute it and/or modify
#       it under the pathways of the GNU General Public License as published by
#       the Free Software Foundation; either version 2 of the License, or
#       (at your option) any later version.
#       
#       This program is distributed in the hope that it will be useful,
#       but WITHOUT ANY WARRANTY; without even the implied warranty of
#       MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
#       GNU General Public License for more details.
#       
#       You should have received a copy of the GNU General Public License
#       along with this program; if not, write to the Free Software
#       Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
#       MA 02110-1301, USA.

import argparse
import os
import sys
from fisher import pvalue as fisher
from decimal import Decimal,getcontext
from math import lgamma,exp,factorial

def binProb(SAPs_KEGG,NoSAPs_KEGG,SAPs_all,NoSAPs_all,CntKEGG_All,totalSAPs,pKEGG):
	"""
	Returns binomial probability.
	"""
	def comb(CntKEGG_All,k):
		return factorial(CntKEGG_All) / Decimal(str(factorial(k)*factorial(CntKEGG_All-k)))
	probLow = 0
	for k in range(0, SAPs_KEGG+1):
		cp=Decimal(str(comb(CntKEGG_All,k)))
		bp=Decimal(str(pKEGG**k))
		dp=Decimal(str(1.0-pKEGG))**Decimal(str(CntKEGG_All-k))
		probLow+=cp*bp*dp
	#~ 
	probHigh = 0
	for k in range(int(SAPs_KEGG),CntKEGG_All+1):
		cp=Decimal(str(comb(CntKEGG_All,k)))
		bp=Decimal(str(pKEGG**k))
		dp=Decimal(str(1.0-pKEGG))**Decimal(str(CntKEGG_All-k))
		probHigh+=cp*bp*dp
	return probLow,probHigh

def gauss_hypergeom(X, CntKEGG_All, SAPs_all, totalSAPs):
	CntKEGG_All,SAPs_all,totalSAPs
	"""
	Returns the probability of drawing X successes of SAPs_all marked items
	in CntKEGG_All draws from a bin of totalSAPs total items
	"""
	def logchoose(ni, ki):
		try:
			lgn1 = lgamma(ni+1)
			lgk1 = lgamma(ki+1)
			lgnk1 = lgamma(ni-ki+1)
		except ValueError:
			raise ValueError
		return lgn1 - (lgnk1 + lgk1)
	#~ 
	r1 = logchoose(SAPs_all, X)
	try:
		r2 = logchoose(totalSAPs-SAPs_all, CntKEGG_All-X)
	except ValueError:
		return 0
	r3 = logchoose(totalSAPs,CntKEGG_All)
	return exp(r1 + r2 - r3)
    
def hypergeo_sf(SAPs_KEGG,NoSAPs_KEGG,SAPs_all,NoSAPs_all,CntKEGG_All,totalSAPs,pKEGG):
	"""
	Runs Hypergeometric probability test
	"""
	s = 0
	t=0
	for i in range(SAPs_KEGG,min(SAPs_all,CntKEGG_All)+1):
		s += max(gauss_hypergeom(i,CntKEGG_All,SAPs_all,totalSAPs), 0.0)
	for i in range(0, SAPs_KEGG+1):
		t += max(gauss_hypergeom(i,CntKEGG_All,SAPs_all,totalSAPs), 0.0)
	return min(max(t,0.0), 1),min(max(s,0.0), 1)

def fisherexct(SAPs_KEGG,NoSAPs_KEGG,SAPs_all,NoSAPs_all,CntKEGG_All,totalSAPs,pKEGG):
	"""
	Runs Fisher's exact test
	"""
	ftest=fisher(SAPs_KEGG,NoSAPs_KEGG,SAPs_all,NoSAPs_all)
	probLow,probHigh=ftest.left_tail,ftest.right_tail
	return probLow,probHigh

def rtrnKEGGcENSEMBLc(inBckgrndfile,columnENSEMBLTBckgrnd,columnKEGGBckgrnd):
	"""
	"""
	dKEGGTENSEMBLT={}
	for eachl in open(inBckgrndfile,'r'):
		if eachl.strip():
			ENSEMBLT=eachl.splitlines()[0].split('\t')[columnENSEMBLTBckgrnd]
			KEGGTs=set(eachl.splitlines()[0].split('\t')[columnKEGGBckgrnd].split('.'))
			KEGGTs=KEGGTs.difference(set(['','U','N']))
			for KEGGT in KEGGTs:
				try:
					dKEGGTENSEMBLT[KEGGT].add(ENSEMBLT)
				except:
					dKEGGTENSEMBLT[KEGGT]=set([ENSEMBLT])
	ENSEMBLTGinKEGG=set.union(*dKEGGTENSEMBLT.values())
	return dKEGGTENSEMBLT,ENSEMBLTGinKEGG

def rtrnENSEMBLcSAPs(inSAPsfile,columnENSEMBLT,ENSEMBLTGinKEGG):
	"""
	returns a set of the ENSEMBLT codes present in the input list and
	in the KEGG file
	"""
	sENSEMBLTSAPsinKEGG=set()
	for eachl in open(inSAPsfile,'r'):
		ENSEMBLT=eachl.splitlines()[0].split('\t')[columnENSEMBLT]
		if ENSEMBLT in ENSEMBLTGinKEGG:
			sENSEMBLTSAPsinKEGG.add(ENSEMBLT)
	return sENSEMBLTSAPsinKEGG

def rtrnCounts(dKEGGTENSEMBLT,ENSEMBLTGinKEGG,sENSEMBLTSAPsinKEGG,statsTest):
	"""
	returns a list of the ENSEMBLT codes present in the input list and
	in the KEGG file. The pathways in this list are: 'Go Term','# Genes in
	the KEGG Term','# Genes in the list and in the KEGG Term','Enrichement
	of the KEGG Term for genes in the input list','Genes in the input list
	present in the KEGG term'
	"""
	totalSAPs=len(ENSEMBLTGinKEGG)
	SAPs_all=len(sENSEMBLTSAPsinKEGG)
	NoSAPs_all=totalSAPs-SAPs_all
	pKEGG=SAPs_all/float(totalSAPs)
	#~ 
	lp=len(dKEGGTENSEMBLT)
	cnt=0
	#~ 
	if statsTest=='fisher':
		ptest=fisherexct
	elif statsTest=='hypergeometric':
		ptest=hypergeo_sf
	elif statsTest=='binomial':
		ptest=binProb
	#~ 
	ltfreqs=[]
	for echKEGGT in dKEGGTENSEMBLT:
		cnt+=1
		CntKEGG_All=len(dKEGGTENSEMBLT[echKEGGT])
		SAPs_KEGG=len(dKEGGTENSEMBLT[echKEGGT].intersection(sENSEMBLTSAPsinKEGG))
		NoSAPs_KEGG=CntKEGG_All-SAPs_KEGG
		probLow,probHigh=ptest(SAPs_KEGG,NoSAPs_KEGG,SAPs_all,NoSAPs_all,CntKEGG_All,totalSAPs,pKEGG)
		ltfreqs.append([(SAPs_KEGG/Decimal(CntKEGG_All)),SAPs_KEGG,probLow,probHigh,echKEGGT])
	#~ 
	ltfreqs.sort()
	ltfreqs.reverse()
	outl=[]
	cper,crank=Decimal('2'),0
	#~ 
	getcontext().prec=2#set 2 decimal places
	for perc,cnt_go,pvalLow,pvalHigh,goTerm in ltfreqs:
		if perc<cper:
			crank+=1
			cper=perc
		outl.append('\t'.join([str(cnt_go),str(Decimal(perc)*Decimal('1.0')),str(crank),str(Decimal(pvalLow)*Decimal('1.0')),str(Decimal(pvalHigh)*Decimal('1.0')),goTerm]))
	#~ 
	return outl
	

def main():
	#~ 
	parser = argparse.ArgumentParser(description='Returns the count of genes in KEGG categories and their statistical overrrepresentation, from a list of genes and an background file (i.e. plane text with ENSEMBLT and KEGG pathways).')
	parser.add_argument('--input',metavar='input TXT file',type=str,help='the input file with the table in txt format.',required=True)
	parser.add_argument('--inBckgrndfile',metavar='input TXT file',type=str,help='the input file with the background table in txt format.',required=True)
	parser.add_argument('--output',metavar='output TXT file',type=str,help='the output file with the table in txt format.',required=True)
	parser.add_argument('--columnENSEMBLT',metavar='column number',type=int,help='column with the ENSEMBL transcript code in the input file.',required=True)
	parser.add_argument('--columnENSEMBLTBckgrnd',metavar='column number',type=int,help='column with the ENSEMBL transcript code in the background file.',required=True)
	parser.add_argument('--columnKEGGBckgrnd',metavar='column number',type=int,help='column with the KEGG pathways in the background file.',required=True)
	parser.add_argument('--statsTest',metavar='input TXT file',type=str,help='statistical test to compare KEGG pathways (i.e. fisher, hypergeometric, binomial).',required=True)

	args = parser.parse_args()

	inSAPsfile = args.input
	inBckgrndfile = args.inBckgrndfile
	saleKEGGPCount = args.output
	columnENSEMBLT = args.columnENSEMBLT
	columnENSEMBLTBckgrnd = args.columnENSEMBLTBckgrnd
	columnKEGGBckgrnd = args.columnKEGGBckgrnd
	statsTest = args.statsTest
	columnENSEMBLT-=1
	columnENSEMBLTBckgrnd-=1
	columnKEGGBckgrnd=-1
	#~ 
	dKEGGTENSEMBLT,ENSEMBLTGinKEGG=rtrnKEGGcENSEMBLc(inBckgrndfile,columnENSEMBLTBckgrnd,columnKEGGBckgrnd)
	sENSEMBLTSAPsinKEGG=rtrnENSEMBLcSAPs(inSAPsfile,columnENSEMBLT,ENSEMBLTGinKEGG)
	outl=rtrnCounts(dKEGGTENSEMBLT,ENSEMBLTGinKEGG,sENSEMBLTSAPsinKEGG,statsTest)
	#~ 
	saleKEGGPCount=open(saleKEGGPCount,'w')
	saleKEGGPCount.write('\n'.join(outl))
	saleKEGGPCount.close()
	#~ 
	return 0

if __name__ == '__main__':
	main()