##################################
#                                #
# Last modified 2025/04/01       # 
#                                #
# Georgi Marinov                 #
#                                # 
##################################

import sys
import string
import math
import pysam
from sets import Set
# import h5py
# import numpy as np    
import os
import re

def FLAG(FLAG):

    Numbers = [0,1,2,4,8,16,32,64,128,256,512,1024,2048]

    FLAGList=[]

    MaxNumberList=[]
    for i in Numbers:
        if i <= FLAG:
            MaxNumberList.append(i)

    Residual=FLAG
    maxPos = len(MaxNumberList)-1

    while Residual > 0:
        if MaxNumberList[maxPos] <= Residual:
            Residual = Residual - MaxNumberList[maxPos]
            FLAGList.append(MaxNumberList[maxPos])
            maxPos-=1
        else:
            maxPos-=1
  
    return FLAGList

def getReverseComplement(preliminarysequence):
    
    DNA = {'A':'T','T':'A','G':'C','C':'G','N':'N','a':'t','t':'a','g':'c','c':'g','n':'n'}
    sequence=''
    for j in range(len(preliminarysequence)):
        sequence=sequence+DNA[preliminarysequence[len(preliminarysequence)-j-1]]
    return sequence

def run():

    if len(sys.argv) < 3:
        print 'usage: python %s bam genome.fa outfile_prefix [-minLen bp] [-excludeChr chr1[,chr2,...,chrN]] [-chrPrefix string] [-strand +|-] [-GAtoCTratio number]' % sys.argv[0]
        sys.exit(1)

    DNA = {'A':'T','T':'A','G':'C','C':'G','N':'N','a':'t','t':'a','g':'c','c':'g','n':'n'}

    doPlusOnly = False
    doMinusOnly = False

    if '-strand' in sys.argv:
        wantedStrand = sys.argv[sys.argv.index('-strand') + 1]
        if wantedStrand == '+':
            doPlusOnly = True 
            print 'will only output plus-strand reads'
        if wantedStrand == '-':
            doMinusOnly = True 
            print 'will only output minus-strand reads'

    doChrPrefix = False
    if '-chrPrefix' in sys.argv:
        doChrPrefix = True
        chrPrefix = sys.argv[sys.argv.index('-chrPrefix') + 1]

    ExcludedChrs = {}
    if '-excludeChr' in sys.argv:
        for chr in sys.argv[sys.argv.index('-excludeChr') + 1].split(','):
            ExcludedChrs[chr] = 1

    minLen = 0
    if '-minLen' in sys.argv:
        minLen = int(sys.argv[sys.argv.index('-excludeChr') + 1])
        print 'will exclude reads shorter than', minLen, 'bp'

    GAtoCTratio = 2
    if '-GAtoCTratio' in sys.argv:
        GAtoCTratio = int(sys.argv[sys.argv.index('-GAtoCTratio') + 1])
        print 'will exclude reads in which the GA/CT conversion ratio is not conclusive; setting it to', GAtoCTratio

    BAM = sys.argv[1]
    fasta = sys.argv[2]
    outprefix = sys.argv[3]

    GenomeDict={}
    sequence=''
    inputdatafile = open(fasta)
    for line in inputdatafile:
        if line[0]=='>':
            if sequence != '':
                GenomeDict[chr] = ''.join(sequence).upper()
            chr = line.strip().split('>')[1]
            print chr, 
            sequence=[]
            Keep=False
            continue
        else:
            sequence.append(line.strip())
    GenomeDict[chr] = ''.join(sequence).upper()

    for chr in ExcludedChrs.keys():
        if GenomeDict.has_key(chr):
           del GenomeDict[chr]

    print 'finished inputting genomic sequence'

    outfile = open(outprefix + '.reads.tsv', 'w')

    samfile = pysam.Samfile(BAM, "rb" )

    InconclusiveReads = 0
    TooShortReads = 0

    RL = 0
    for read in samfile.fetch(until_eof=True):
        RL+=1
        if RL % 100000 == 0:
            print str(RL/100000) + 'M alignments processed processed', chr
        fields = str(read).split('\t')
        ID = read.qname
        if read.is_unmapped:
            continue
        chr = samfile.getrname(read.tid)
        if GenomeDict.has_key(chr):
            pass
        else:
            continue
        FLAGfields = FLAG(int(fields[1]))
        if 256 in FLAGfields:
            continue
        if 16 in FLAGfields:
            strand = '-'
            if doPlusOnly:
                continue
            sequence = fields[9]
        else:
            strand = '+'
            if doMinusOnly:
                continue
            sequence = fields[9]
        if len(sequence) < minLen:
            TooShortReads += 1
            continue

        pos = int(fields[3])
        
        outputPositions = []
        outputProbs = []

        ReadPosToGenomePos = {}
        GenomePosToReadPos = {}
        currentPos = pos
        readPos = 0
        for (m,bp) in read.cigar:
            if m == 0:
                for j in range(bp):
                    ReadPosToGenomePos[readPos + j] = currentPos + j
                    GenomePosToReadPos[currentPos + j] = readPos + j
                readPos += bp
                currentPos += bp
            elif m == 5:
                pass
            elif m == 4:
                readPos += bp
            elif m == 2:
                currentPos += bp
            elif m == 3:
                currentPos += bp
            elif m == 1:
                readPos += bp
            else:
                print ID,( m,bp), 'unrecognized CIGAR field, exiting'
                sys.exit(1)

        MMDict = {}
        MMDict[('G','A')] = 0.01
        MMDict[('C','T')] = 0.01

        for pos in ReadPosToGenomePos.keys():
            Rseq = sequence[pos]
            Gseq = GenomeDict[chr][ReadPosToGenomePos[pos]]
            if Rseq != Gseq:
                MM = (Gseq,Rseq)
                if MMDict.has_key(MM):
                    pass
                else:
                    MMDict[MM] = 0.0
                MMDict[MM] += 1

        print 'GA:', MMDict[('G','A')], 'CT:', MMDict[('C','T')]

        if MMDict[('G','A')]/MMDict[('C','T')] > GAtoCTratio:
            Conversion = 'GA'
        elif MMDict[('C','T')]/MMDict[('G','A')] > GAtoCTratio:
            Conversion = 'CT'
        else:
            InconclusiveReads += 1
            continue

        start = min(GenomePosToReadPos.keys())
        end = max(GenomePosToReadPos.keys())

        print start, end

        if Conversion == 'CT':
            positions = [m.start() + start for m in re.finditer('C', GenomeDict[chr][start:end])]
            for pos in positions:
                try:
                    Rseq = sequence[GenomePosToReadPos[pos]]
                    Gseq = GenomeDict[chr][pos]
                    if Gseq == 'C' and Rseq == 'T': 
                        outputProbs += [1]
                    else:
                        outputProbs += [0]
                except:
                    outputProbs += [0]
        if Conversion == 'GA':
            positions = [m.start() + start for m in re.finditer('G', GenomeDict[chr][start:end])]
            for pos in positions:
                try:
                    Rseq = sequence[GenomePosToReadPos[pos]]
                    Gseq = GenomeDict[chr][pos]
                    if Gseq == 'G' and Rseq == 'A': 
                        outputProbs += [1]
                    else:
                        outputProbs += [0]
                except:
                    outputProbs += [0]

        preliminaryOutput = zip(positions, outputProbs)
        preliminaryOutput.sort()

        outline = chr + '\t' + str(start) + '\t' + str(end) + '\t' + strand + '\t' + ID + '\t' + '.'
        Ps = ''
        LLs = ''
        for (gpos,ll) in preliminaryOutput:
            Ps = Ps + str(gpos) + ','
            LLs = LLs + str(ll) + ','
        outline = outline + '\t' + Ps[0:-1]
        outline = outline + '\t' + LLs[0:-1]
        outfile.write(outline + '\n')

    outfile.close()

    
run()