CowChrom <- c(as.character(seq(1 : 29)), "X", "Y")

min.align.len <- 49


##Preparataion of Megablast output table (D) and of position table (POSILMN)
##The Megablast output table processed such that fields will be atomic
##The position table is filtered (Build information is discarded)

if(1)
{
	#Reading in file containing the chromosomal positions as detemined by Illumina
	POSORIG <- read.table(file = "../Material/BovineSNP50_Final_SNPs_54001.csv", header = T, sep = ",")
	names(POSORIG) <- c("ID", "BUILD", "CHROM", "POS")
	
	#Reading in file containing the megablast output
	D <- read.table(file = "../Material/OligosVsBovineGenome.megablast", header = F)
	names(D) <- c("QUERY", "SUBJECT", "IDENT", "ALIGN.LEN", "MISMATCHES",
	"GAPS", "Q.START", "Q.END", "S.START", "S.END", 
	"E.VALUE", "BIT.SCORE")
	

	cat("Number of Hits:\t", nrow(D), "\n")
		
	#stop()	
	D <- subset(D, ALIGN.LEN >= min.align.len)	
	
	n <- nrow(D)
	cat("Number of hits with an alignment length of ", min.align.len, " or greater:\t", n, "\n")
	
	
	
	
	
	#Pulling out the chromosome from the column SUBJECT (i.e. the hit sequence or genome sequence)	
	Q <- strsplit(as.character(D$SUBJECT), ":")	
	CHROM <- vector(mode = "character", length = n)
	for(i in 1 : n)
	{
		CHROM[i] <- Q[[i]][3]
	}	
	SNPID        <- vector(mode = "character", length = n)
	ILMNSTRAND   <- vector(mode = "character", length = n)
	SOURCESTRAND <- vector(mode = "character", length = n)

	
	
	#Pulling out the SNPID + strand information from the column QUERY (i.e. the oligo sequence or search sequence)
	Q <- strsplit(as.character(D$QUERY), "|", fixed = TRUE)
	for(i in 1 : n)
	{
		SNPID[i] <- Q[[i]][1]
		ILMNSTRAND[i] <- Q[[i]][2]
		SOURCESTRAND[i] <- Q[[i]][3]
	}
	D <- cbind(D, SNPID, CHROM, ILMNSTRAND, SOURCESTRAND)	
	#stop()
	D <- subset(D, CHROM %in% CowChrom)	
	n <- nrow(D)
	cat("Number of hits on autosome or X chromosome:\t", n, "\n")
			
	POSILMN <- subset(POSORIG, select = c("CHROM", "ID", "POS"))			
}

#stop()

l <- nrow(POSILMN)
n.multiplehit <- 0
n.nohit <- 0
n.uniquehit <- 0

n.method1 <- 0
n.method2 <- 0
n.method3 <- 0

CHROM <- vector(mode = "character", length = l)
POS <- vector(mode = "numeric", length = l)

for(i in 1 : l)
{
	id <- as.character(POSILMN$ID[i])	
	#cat("Number:\t", i, "SNP:\t", id, "\n")	
	D.aux <- subset(D, SNPID == id)
	
	#stop()
	
	OK <- TRUE
	
	### Just one hit
	if(nrow(D.aux) == 1)
	{
		n.uniquehit <- n.uniquehit + 1
		#cat("\teindeutig\n")
					
		CHROM[i] <- as.character(D.aux$CHROM[1])											
		if(D.aux$S.START < D.aux$S.END)
		{
			POS[i] <- D.aux$S.END[1] + 1
		}
		else
		{
			POS[i] <- D.aux$S.END[1] - 1
		}
		OK <- TRUE			
	}
								
	### No hit
	if(nrow(D.aux) == 0)
	{
		n.nohit <- n.nohit + 1
		#cat("\tNo hit\n")
		is.na(CHROM[i]) <- TRUE
		is.na(POS[i]) <- TRUE
		OK <- TRUE
	}
	### More than one hit	
	if(nrow(D.aux) > 1)
	{	
		#cat("\tmultiplehit\n")
		n.multiplehit <- n.multiplehit + 1
		OK <- FALSE
	}
	#stop()
	
	###########################################################
	
	##### D I S A M B I G U A T I O N  ########################
	
	###########################################################	
	
	### Method 1
	### Pull out hit with the best bit score
	### If there is just one hit with the best bit score we are done
		
	if(nrow(D.aux) > 1 && OK == FALSE)
	{
		
		#cat("Methode 1 - multiple hits:\n")
		#print(D.aux)		
		
		#cat("Original Position:\n")
		#print(subset(POSILMN, ID == id))
								
		max.bitscore <- max(D.aux$BIT.SCORE)
		
		D.aux <- subset(D.aux, BIT.SCORE == max.bitscore)
		#print(D.aux)
		if(nrow(D.aux) == 1)
		{			
			CHROM[i] <- as.character(D.aux$CHROM[1])									
			if(D.aux$S.START < D.aux$S.END)
			{
				POS[i] <- D.aux$S.END[1] + 1
			}
			else
			{
				POS[i] <- D.aux$S.END[1] - 1
			}
			OK <- TRUE			
			n.method1 <- n.method1 + 1
			#cat("Method 1 successful\n")
		}
		else #Two or more equally good hits
		{
			chroms <- unique(D.aux$CHROM)		
			if(length(chroms) == 1 && length(unique(D.aux$S.START)) > 1) #One chromosome but several different positions
			{
				CHROM[i] <- as.character(unique(D.aux$CHROM))
				is.na(POS[i]) <- TRUE			
				OK <- TRUE
				n.method2 <- n.method2 + 1
				#cat("Method 2 successful\n")
			}	
			else #Several chromsomes
			{
				is.na(POS[i]) <- TRUE
				is.na(CHROM[i]) <- TRUE
				n.method3 <- n.method3 + 1
			}													
		}											
	}						
}

CHROM.MAP <- CHROM
POS.MAP <- POS

POSILMN <- cbind(POSILMN, CHROM.MAP, POS.MAP)

names(POSILMN) <- c("CHROM.ILMN", "ID", "POS.ILMN", "CHROM.MAP", "POS.MAP")

POSILMN <- subset(POSILMN, select = c("ID", "CHROM.ILMN", "POS.ILMN", "CHROM.MAP", "POS.MAP"))

write.table(POSILMN, file = "Mapping-Illumina-Genome-25-June-2009.txt", quote = F, row.names = F, sep = "\t")


cat("Number of multiple hits:\t", n.multiplehit, "\n")
cat("Number of oligos without hit:\t", n.nohit, "\n")
cat("Number of oligos with unique hit:\t", n.uniquehit, "\n")

cat("Disambiguation\n")
cat("\tMethod 1:\t", n.method1, "\n")
cat("\tMethod 2:\t", n.method2, "\n")
cat("\tMethod 3:\t", n.method3, "\n")