# Written by Jay Taylor (3/7/2018)
# Modified by Ti Eriksson (last 11/29/2018)
###################################################################
# Relatedness statistics for haplodiploid colonies (7 March 2018)

# This code generates random samples of worker genotypes assuming different degrees of 
# relatedness between the parents.
# Four unlinked microsatellite loci are simulated, using allele frequencies supplied
# by the user.
# Scenarios simulated: I and II


# Load packages and function files

library(related)
source("functions.r")


# Defined parameters

Nsim = 1000	# number of simulations used to estimate each p-value
Nloci = 4	# number of microsatellite loci
mf = 2		# number of male mates per queen

# Read in allele frequencies

freqs <- read.table("allele_freqs.txt", header=TRUE, sep="\t")
locus1 <- subset(freqs, locus == 1, select=c(allele, freq))
locus2 <- subset(freqs, locus == 2, select=c(allele, freq))
locus3 <- subset(freqs, locus == 3, select=c(allele, freq))
locus4 <- subset(freqs, locus == 4, select=c(allele, freq))
allele_list <- list(locus1,locus2,locus3,locus4)

# Read in colony relatedness statistics

cdata <- read.table("colonies.txt", header=TRUE, sep="\t")

Ncolonies <- length(cdata$colony)	# number of colonies in data set


# Relatedness p-values

Pval1 = numeric(Ncolonies)
Pval2 = numeric(Ncolonies)

for (col in 1:Ncolonies) {

  Nworkers <- cdata$n[col]
  Nqueens <- cdata$q[col]
  Ndrones <- mf*Nqueens
  meanR <- cdata$R[col] 	# mean pairwise R for this colony
  names = c(1:Nworkers)
 
  ########################################################################################################
  # Scenario I: q unrelated queens each mated to mf unrelated fathers

  meanRsim1 <- numeric(Nsim)	# This array will contain the simulated mean pairwise R's

  for (sim in 1:Nsim)  {
    
    # Sample the unrelated queens and drones
 
    queen <- matrix(0,Nqueens,2*Nloci)
    for (q in 1:Nqueens)  {
      genotype <- diploid_genotype(Nloci, allele_list)
      queen[q,1:(2*Nloci)] <- t(genotype)
    }

    drone <- matrix(0,Ndrones,Nloci)
    for (d in 1:Ndrones)  {
      genotype <- haploid_genotype(Nloci, allele_list)
      drone[d,1:Nloci] <- t(genotype)
    }

    # Sample the offspring (worker) genotypes
    worker <- matrix(0,Nworkers,2*Nloci)
    for (w in 1:Nworkers) {
      q <- runif(1,1,Nqueens)  		# choose the mother
      d <- runif(1,mf*(q-1)+1,mf*q)  	# choose the father
      qg <- queen[q,1:(2*Nloci)]
      dg <- drone[d,1:Nloci]
      genotype <- offspring_genotype(Nloci, qg, dg)
      worker[w,1:(2*Nloci)] <- t(genotype)
    }   

    # Calculate pairwise relatedness
    gdata <- data.frame(names,worker)	# data frame with worker names and genotypes
    output1 <- coancestry(gdata, allele.freqs=allele_list, quellergt=1)
    meanRsim1[sim] <- mean(output1$relatedness[[10]])	# calculate and save the mean pairwise R
    
  }

  # Calculate the p-value of the observed mean pairwise R for this colony relative to the simulated values
  Pval1[col] <- (rank(c(meanR,meanRsim1))[1])/(Nsim + 1)

  ########################################################################################################



  ########################################################################################################
  # Scenario II: q queens that all share a common mother and mf unrelated fathers 

  meanRsim2 <- numeric(Nsim)	# This array will contain the simulated mean pairwise R's

  for (sim in 1:Nsim)  {
    

    # Sample the grand-queen and her male mates

    grandqueen <- diploid_genotype(Nloci, allele_list)
    granddrone <- matrix(0,mf,Nloci)
    for (d in 1:mf)  {
      genotype <- haploid_genotype(Nloci, allele_list)
      granddrone[d,1:Nloci] <- t(genotype)
    }

    # Sample the queens and their unrelated male mates
 
    queen <- matrix(0,Nqueens,2*Nloci)
    for (q in 1:Nqueens)  {
      qg <- grandqueen
      d <- runif(1,1,mf)  	# choose the father (granddrone)
      dg <- granddrone[d,1:Nloci]
      genotype <- offspring_genotype(Nloci, qg, dg)
      queen[q,1:(2*Nloci)] <- t(genotype)
    }

    drone <- matrix(0,Ndrones,Nloci)
    for (d in 1:Ndrones)  {
      genotype <- haploid_genotype(Nloci, allele_list)
      drone[d,1:Nloci] <- t(genotype)
    }

    # Sample the offspring (worker) genotypes
    worker <- matrix(0,Nworkers,2*Nloci)
    for (w in 1:Nworkers) {
      q <- runif(1,1,Nqueens)  		# choose the mother
      d <- runif(1,mf*(q-1)+1,mf*q)  	# choose the father
      qg <- queen[q,1:(2*Nloci)]
      dg <- drone[d,1:Nloci]
      genotype <- offspring_genotype(Nloci, qg, dg)
      worker[w,1:(2*Nloci)] <- t(genotype)
    }   

    # Calculate pairwise relatedness
    gdata <- data.frame(names,worker)	# data frame with worker names and genotypes
    output2 <- coancestry(gdata, allele.freqs=allele_list, quellergt=1)
    meanRsim2[sim] <- mean(output2$relatedness[[10]])	# calculate and save the mean pairwise R
  }

  # Calculate the p-value of the observed mean pairwise R for this colony relative to the simulated values
  Pval2[col] <- (rank(c(meanR,meanRsim2))[1])/(Nsim + 1)
  ########################################################################################################
}
as.data.frame(Pval1)
as.data.frame(Pval2)




########################################################################################################
# Relatedness statistics for haplodiploid colonies (7 March 2018)

# This code generates random samples of worker genotypes assuming different degrees of 
# relatedness between the parents.
# Four unlinked microsatellite loci are simulated, using allele frequencies supplied
# by the user.
# Scenarios simulated: III-VIII

# Load packages and function files

library(related)
source("functions.r")


# Defined parameters

Nsim = 1000	# number of simulations used to estimate each p-value
Nloci = 4	# number of microsatellite loci
mf = 2		# number of male mates per queen
grandqueen2<-matrix(ncol=8,nrow=2)
grandqueen3<-matrix(ncol=8,nrow=3)
grandqueen4<-matrix(ncol=8,nrow=4)
grandqueen5<-matrix(ncol=8,nrow=5)
grandqueen6<-matrix(ncol=8,nrow=6)
grandqueen7<-matrix(ncol=8,nrow=7)
grandqueen8<-matrix(ncol=8,nrow=8)

# Read in allele frequencies

freqs <- read.table("allele_freqs.txt", header=TRUE, sep="\t")
locus1 <- subset(freqs, locus == 1, select=c(allele, freq))
locus2 <- subset(freqs, locus == 2, select=c(allele, freq))
locus3 <- subset(freqs, locus == 3, select=c(allele, freq))
locus4 <- subset(freqs, locus == 4, select=c(allele, freq))
allele_list <- list(locus1,locus2,locus3,locus4)


# Read in colony relatedness statistics

cdata <- read.table("colonies.txt", header=TRUE, sep="\t")

Ncolonies <- length(cdata$colony)	# number of colonies in data set


# Relatedness p-values

Pval3 = numeric(Ncolonies)
Pval4 = numeric(Ncolonies)
Pval5 = numeric(Ncolonies)
Pval6 = numeric(Ncolonies)
Pval7 = numeric(Ncolonies)
Pval8 = numeric(Ncolonies)
Pval9 = numeric(Ncolonies)

for (col in 1:Ncolonies) {

  Nworkers <- cdata$n[col]
  Nqueens <- cdata$q[col]
  Ndrones <- mf*Nqueens
  meanR <- cdata$R[col] 	# mean pairwise R for this colony
  names = c(1:Nworkers)
 

  ########################################################################################################
  # Scenario III: q queens that share 2 common mothers and mf unrelated fathers 

  meanRsim2 <- numeric(Nsim)	# This array will contain the simulated mean pairwise R's

  for (sim in 1:Nsim)  {
    

    # Sample the two grand-queens and their male mates

    grandqueen2[1,1:(2*Nloci)] <- diploid_genotype(Nloci, allele_list)
    grandqueen2[2,1:(2*Nloci)] <- diploid_genotype(Nloci, allele_list)
    granddrone <- matrix(0,2*mf,Nloci)
    for (d in 1:(2*mf))  {
      genotype <- haploid_genotype(Nloci, allele_list)
      granddrone[d,1:Nloci] <- t(genotype)
    }

    # Sample the queens and their unrelated male mates
 
    queen <- matrix(0,Nqueens,2*Nloci)
    for (q in 1:Nqueens)  {
      g <- runif(1,1,2)		# choose the mother (grandquuen)
      qg <- grandqueen2[g,1:(2*Nloci)]
      d <- runif(1,1,mf)  	# choose the father (granddrone)
      dg <- granddrone[(g-1)*mf+d,1:Nloci]
      genotype <- offspring_genotype(Nloci, qg, dg)
      queen[q,1:(2*Nloci)] <- t(genotype)
    }

    drone <- matrix(0,Ndrones,Nloci)
    for (d in 1:Ndrones)  {
      genotype <- haploid_genotype(Nloci, allele_list)
      drone[d,1:Nloci] <- t(genotype)
    }

    # Sample the offspring (worker) genotypes
    worker <- matrix(0,Nworkers,2*Nloci)
    for (w in 1:Nworkers) {
      q <- runif(1,1,Nqueens)  		# choose the mother
      d <- runif(1,mf*(q-1)+1,mf*q)  	# choose the father
      qg <- queen[q,1:(2*Nloci)]
      dg <- drone[d,1:Nloci]
      genotype <- offspring_genotype(Nloci, qg, dg)
      worker[w,1:(2*Nloci)] <- t(genotype)
    }   

    # Calculate pairwise relatedness
    gdata <- data.frame(names,worker)	# data frame with worker names and genotypes
    output <- coancestry(gdata, allele.freqs=allele_list, quellergt=1)
    meanRsim2[sim] <- mean(output$relatedness[[10]])	# calculate and save the mean pairwise R
    
  }

  # Calculate the p-value of the observed mean pairwise R for this colony relative to the simulated values
  Pval3[col] <- (rank(c(meanR,meanRsim2))[1])/(Nsim + 1)

  ########################################################################################################


  ########################################################################################################
  # Scenario 4: q queens that share 3 common mothers and mf unrelated fathers 

  meanRsim3 <- numeric(Nsim)	# This array will contain the simulated mean pairwise R's

  for (sim in 1:Nsim)  {
    

    # Sample the two grand-queens and their male mates

    grandqueen3[1,1:(2*Nloci)] <- diploid_genotype(Nloci, allele_list)
    grandqueen3[2,1:(2*Nloci)] <- diploid_genotype(Nloci, allele_list)
    grandqueen3[3,1:(2*Nloci)] <- diploid_genotype(Nloci, allele_list)
    granddrone <- matrix(0,3*mf,Nloci)
    for (d in 1:(3*mf))  {
      genotype <- haploid_genotype(Nloci, allele_list)
      granddrone[d,1:Nloci] <- t(genotype)
    }

    # Sample the queens and their unrelated male mates
 
    queen <- matrix(0,Nqueens,2*Nloci)
    for (q in 1:Nqueens)  {
      g <- runif(1,1,3)		# choose grandqueen
      qg <- grandqueen3[g,1:(2*Nloci)]
      d <- runif(1,1,mf)  	# choose granddrone
      dg <- granddrone[(g-1)*mf+d,1:Nloci]
      genotype <- offspring_genotype(Nloci, qg, dg)
      queen[q,1:(2*Nloci)] <- t(genotype)
    }

    drone <- matrix(0,Ndrones,Nloci)
    for (d in 1:Ndrones)  {
      genotype <- haploid_genotype(Nloci, allele_list)
      drone[d,1:Nloci] <- t(genotype)
    }

    # Sample the offspring (worker) genotypes
    worker <- matrix(0,Nworkers,2*Nloci)
    for (w in 1:Nworkers) {
      q <- runif(1,1,Nqueens)  		# choose the mother
      d <- runif(1,mf*(q-1)+1,mf*q)  	# choose the father
      qg <- queen[q,1:(2*Nloci)]
      dg <- drone[d,1:Nloci]
      genotype <- offspring_genotype(Nloci, qg, dg)
      worker[w,1:(2*Nloci)] <- t(genotype)
    }   

    # Calculate pairwise relatedness
    gdata <- data.frame(names,worker)	# data frame with worker names and genotypes
    output <- coancestry(gdata, allele.freqs=allele_list, quellergt=1)
    meanRsim3[sim] <- mean(output$relatedness[[10]])	# calculate and save the mean pairwise R
    
  }

  # Calculate the p-value of the observed mean pairwise R for this colony relative to the simulated values
  Pval4[col] <- (rank(c(meanR,meanRsim3))[1])/(Nsim + 1)

  ########################################################################################################

 ########################################################################################################
  # Scenario 5: q queens that share 4 common mothers and mf unrelated fathers 

  meanRsim4 <- numeric(Nsim)	# This array will contain the simulated mean pairwise R's

  for (sim in 1:Nsim)  {
    

    # Sample the two grand-queens and their male mates

    grandqueen4[1,1:(2*Nloci)] <- diploid_genotype(Nloci, allele_list)
    grandqueen4[2,1:(2*Nloci)] <- diploid_genotype(Nloci, allele_list)
    grandqueen4[3,1:(2*Nloci)] <- diploid_genotype(Nloci, allele_list)
    grandqueen4[4,1:(2*Nloci)] <- diploid_genotype(Nloci, allele_list)

    granddrone <- matrix(0,4*mf,Nloci)
    for (d in 1:(4*mf))  {
      genotype <- haploid_genotype(Nloci, allele_list)
      granddrone[d,1:Nloci] <- t(genotype)
    }

    # Sample the queens and their unrelated male mates
 
    queen <- matrix(0,Nqueens,2*Nloci)
    for (q in 1:Nqueens)  {
      g <- runif(1,1,3)		# choose grandqueen
      qg <- grandqueen4[g,1:(2*Nloci)]
      d <- runif(1,1,mf)  	# choose granddrone
      dg <- granddrone[(g-1)*mf+d,1:Nloci]
      genotype <- offspring_genotype(Nloci, qg, dg)
      queen[q,1:(2*Nloci)] <- t(genotype)
    }

    drone <- matrix(0,Ndrones,Nloci)
    for (d in 1:Ndrones)  {
      genotype <- haploid_genotype(Nloci, allele_list)
      drone[d,1:Nloci] <- t(genotype)
    }

    # Sample the offspring (worker) genotypes
    worker <- matrix(0,Nworkers,2*Nloci)
    for (w in 1:Nworkers) {
      q <- runif(1,1,Nqueens)  		# choose the mother
      d <- runif(1,mf*(q-1)+1,mf*q)  	# choose the father
      qg <- queen[q,1:(2*Nloci)]
      dg <- drone[d,1:Nloci]
      genotype <- offspring_genotype(Nloci, qg, dg)
      worker[w,1:(2*Nloci)] <- t(genotype)
    }   

    # Calculate pairwise relatedness
    gdata <- data.frame(names,worker)	# data frame with worker names and genotypes
    output <- coancestry(gdata, allele.freqs=allele_list, quellergt=1)
    meanRsim4[sim] <- mean(output$relatedness[[10]])	# calculate and save the mean pairwise R
    
  }

  # Calculate the p-value of the observed mean pairwise R for this colony relative to the simulated values
  Pval5[col] <- (rank(c(meanR,meanRsim4))[1])/(Nsim + 1)

  ########################################################################################################

 ########################################################################################################
  # Scenario 6: q queens that share 5 common mothers and mf unrelated fathers 

  meanRsim5 <- numeric(Nsim)	# This array will contain the simulated mean pairwise R's

  for (sim in 1:Nsim)  {
    

    # Sample the two grand-queens and their male mates

    grandqueen5[1,1:(2*Nloci)] <- diploid_genotype(Nloci, allele_list)
    grandqueen5[2,1:(2*Nloci)] <- diploid_genotype(Nloci, allele_list)
    grandqueen5[3,1:(2*Nloci)] <- diploid_genotype(Nloci, allele_list)
    grandqueen5[4,1:(2*Nloci)] <- diploid_genotype(Nloci, allele_list)
    grandqueen5[5,1:(2*Nloci)] <- diploid_genotype(Nloci, allele_list)


    granddrone <- matrix(0,5*mf,Nloci)
    for (d in 1:(5*mf))  {
      genotype <- haploid_genotype(Nloci, allele_list)
      granddrone[d,1:Nloci] <- t(genotype)
    }

    # Sample the queens and their unrelated male mates
 
    queen <- matrix(0,Nqueens,2*Nloci)
    for (q in 1:Nqueens)  {
      g <- runif(1,1,3)		# choose grandqueen
      qg <- grandqueen5[g,1:(2*Nloci)]
      d <- runif(1,1,mf)  	# choose granddrone
      dg <- granddrone[(g-1)*mf+d,1:Nloci]
      genotype <- offspring_genotype(Nloci, qg, dg)
      queen[q,1:(2*Nloci)] <- t(genotype)
    }

    drone <- matrix(0,Ndrones,Nloci)
    for (d in 1:Ndrones)  {
      genotype <- haploid_genotype(Nloci, allele_list)
      drone[d,1:Nloci] <- t(genotype)
    }

    # Sample the offspring (worker) genotypes
    worker <- matrix(0,Nworkers,2*Nloci)
    for (w in 1:Nworkers) {
      q <- runif(1,1,Nqueens)  		# choose the mother
      d <- runif(1,mf*(q-1)+1,mf*q)  	# choose the father
      qg <- queen[q,1:(2*Nloci)]
      dg <- drone[d,1:Nloci]
      genotype <- offspring_genotype(Nloci, qg, dg)
      worker[w,1:(2*Nloci)] <- t(genotype)
    }   

    # Calculate pairwise relatedness
    gdata <- data.frame(names,worker)	# data frame with worker names and genotypes
    output <- coancestry(gdata, allele.freqs=allele_list, quellergt=1)
    meanRsim5[sim] <- mean(output$relatedness[[10]])	# calculate and save the mean pairwise R
    
  }

  # Calculate the p-value of the observed mean pairwise R for this colony relative to the simulated values
  Pval6[col] <- (rank(c(meanR,meanRsim5))[1])/(Nsim + 1)

  ########################################################################################################

########################################################################################################
  # Scenario 7: q queens that share 6 common mothers and mf unrelated fathers 

  meanRsim6 <- numeric(Nsim)	# This array will contain the simulated mean pairwise R's

  for (sim in 1:Nsim)  {
    

    # Sample the two grand-queens and their male mates

    grandqueen6[1,1:(2*Nloci)] <- diploid_genotype(Nloci, allele_list)
    grandqueen6[2,1:(2*Nloci)] <- diploid_genotype(Nloci, allele_list)
    grandqueen6[3,1:(2*Nloci)] <- diploid_genotype(Nloci, allele_list)
    grandqueen6[4,1:(2*Nloci)] <- diploid_genotype(Nloci, allele_list)
    grandqueen6[5,1:(2*Nloci)] <- diploid_genotype(Nloci, allele_list)
    grandqueen6[6,1:(2*Nloci)] <- diploid_genotype(Nloci, allele_list)


    granddrone <- matrix(0,6*mf,Nloci)
    for (d in 1:(6*mf))  {
      genotype <- haploid_genotype(Nloci, allele_list)
      granddrone[d,1:Nloci] <- t(genotype)
    }

    # Sample the queens and their unrelated male mates
 
    queen <- matrix(0,Nqueens,2*Nloci)
    for (q in 1:Nqueens)  {
      g <- runif(1,1,3)		# choose grandqueen
      qg <- grandqueen6[g,1:(2*Nloci)]
      d <- runif(1,1,mf)  	# choose granddrone
      dg <- granddrone[(g-1)*mf+d,1:Nloci]
      genotype <- offspring_genotype(Nloci, qg, dg)
      queen[q,1:(2*Nloci)] <- t(genotype)
    }

    drone <- matrix(0,Ndrones,Nloci)
    for (d in 1:Ndrones)  {
      genotype <- haploid_genotype(Nloci, allele_list)
      drone[d,1:Nloci] <- t(genotype)
    }

    # Sample the offspring (worker) genotypes
    worker <- matrix(0,Nworkers,2*Nloci)
    for (w in 1:Nworkers) {
      q <- runif(1,1,Nqueens)  		# choose the mother
      d <- runif(1,mf*(q-1)+1,mf*q)  	# choose the father
      qg <- queen[q,1:(2*Nloci)]
      dg <- drone[d,1:Nloci]
      genotype <- offspring_genotype(Nloci, qg, dg)
      worker[w,1:(2*Nloci)] <- t(genotype)
    }   

    # Calculate pairwise relatedness
    gdata <- data.frame(names,worker)	# data frame with worker names and genotypes
    output <- coancestry(gdata, allele.freqs=allele_list, quellergt=1)
    meanRsim6[sim] <- mean(output$relatedness[[10]])	# calculate and save the mean pairwise R
    
  }

  # Calculate the p-value of the observed mean pairwise R for this colony relative to the simulated values
  Pval7[col] <- (rank(c(meanR,meanRsim6))[1])/(Nsim + 1)

  ########################################################################################################

########################################################################################################
  # Scenario 8: q queens that share 7 common mothers and mf unrelated fathers 

  meanRsim7 <- numeric(Nsim)	# This array will contain the simulated mean pairwise R's

  for (sim in 1:Nsim)  {
    

    # Sample the two grand-queens and their male mates

    grandqueen7[1,1:(2*Nloci)] <- diploid_genotype(Nloci, allele_list)
    grandqueen7[2,1:(2*Nloci)] <- diploid_genotype(Nloci, allele_list)
    grandqueen7[3,1:(2*Nloci)] <- diploid_genotype(Nloci, allele_list)
    grandqueen7[4,1:(2*Nloci)] <- diploid_genotype(Nloci, allele_list)
    grandqueen7[5,1:(2*Nloci)] <- diploid_genotype(Nloci, allele_list)
    grandqueen7[6,1:(2*Nloci)] <- diploid_genotype(Nloci, allele_list)
    grandqueen7[7,1:(2*Nloci)] <- diploid_genotype(Nloci, allele_list)


    granddrone <- matrix(0,7*mf,Nloci)
    for (d in 1:(7*mf))  {
      genotype <- haploid_genotype(Nloci, allele_list)
      granddrone[d,1:Nloci] <- t(genotype)
    }

    # Sample the queens and their unrelated male mates
 
    queen <- matrix(0,Nqueens,2*Nloci)
    for (q in 1:Nqueens)  {
      g <- runif(1,1,3)		# choose grandqueen
      qg <- grandqueen7[g,1:(2*Nloci)]
      d <- runif(1,1,mf)  	# choose granddrone
      dg <- granddrone[(g-1)*mf+d,1:Nloci]
      genotype <- offspring_genotype(Nloci, qg, dg)
      queen[q,1:(2*Nloci)] <- t(genotype)
    }

    drone <- matrix(0,Ndrones,Nloci)
    for (d in 1:Ndrones)  {
      genotype <- haploid_genotype(Nloci, allele_list)
      drone[d,1:Nloci] <- t(genotype)
    }

    # Sample the offspring (worker) genotypes
    worker <- matrix(0,Nworkers,2*Nloci)
    for (w in 1:Nworkers) {
      q <- runif(1,1,Nqueens)  		# choose the mother
      d <- runif(1,mf*(q-1)+1,mf*q)  	# choose the father
      qg <- queen[q,1:(2*Nloci)]
      dg <- drone[d,1:Nloci]
      genotype <- offspring_genotype(Nloci, qg, dg)
      worker[w,1:(2*Nloci)] <- t(genotype)
    }   

    # Calculate pairwise relatedness
    gdata <- data.frame(names,worker)	# data frame with worker names and genotypes
    output <- coancestry(gdata, allele.freqs=allele_list, quellergt=1)
    meanRsim7[sim] <- mean(output$relatedness[[10]])	# calculate and save the mean pairwise R
    
  }

  # Calculate the p-value of the observed mean pairwise R for this colony relative to the simulated values
  Pval8[col] <- (rank(c(meanR,meanRsim7))[1])/(Nsim + 1)

  ########################################################################################################

########################################################################################################
  # Scenario 9: q queens that share 8 common mothers and mf unrelated fathers 

  meanRsim8 <- numeric(Nsim)	# This array will contain the simulated mean pairwise R's

  for (sim in 1:Nsim)  {
    

    # Sample the two grand-queens and their male mates

    grandqueen8[1,1:(2*Nloci)] <- diploid_genotype(Nloci, allele_list)
    grandqueen8[2,1:(2*Nloci)] <- diploid_genotype(Nloci, allele_list)
    grandqueen8[3,1:(2*Nloci)] <- diploid_genotype(Nloci, allele_list)
    grandqueen8[4,1:(2*Nloci)] <- diploid_genotype(Nloci, allele_list)
    grandqueen7[5,1:(2*Nloci)] <- diploid_genotype(Nloci, allele_list)
    grandqueen8[6,1:(2*Nloci)] <- diploid_genotype(Nloci, allele_list)
    grandqueen8[7,1:(2*Nloci)] <- diploid_genotype(Nloci, allele_list)
    grandqueen8[8,1:(2*Nloci)] <- diploid_genotype(Nloci, allele_list)


    granddrone <- matrix(0,8*mf,Nloci)
    for (d in 1:(8*mf))  {
      genotype <- haploid_genotype(Nloci, allele_list)
      granddrone[d,1:Nloci] <- t(genotype)
    }

    # Sample the queens and their unrelated male mates
 
    queen <- matrix(0,Nqueens,2*Nloci)
    for (q in 1:Nqueens)  {
      g <- runif(1,1,3)		# choose grandqueen
      qg <- grandqueen8[g,1:(2*Nloci)]
      d <- runif(1,1,mf)  	# choose granddrone
      dg <- granddrone[(g-1)*mf+d,1:Nloci]
      genotype <- offspring_genotype(Nloci, qg, dg)
      queen[q,1:(2*Nloci)] <- t(genotype)
    }

    drone <- matrix(0,Ndrones,Nloci)
    for (d in 1:Ndrones)  {
      genotype <- haploid_genotype(Nloci, allele_list)
      drone[d,1:Nloci] <- t(genotype)
    }

    # Sample the offspring (worker) genotypes
    worker <- matrix(0,Nworkers,2*Nloci)
    for (w in 1:Nworkers) {
      q <- runif(1,1,Nqueens)  		# choose the mother
      d <- runif(1,mf*(q-1)+1,mf*q)  	# choose the father
      qg <- queen[q,1:(2*Nloci)]
      dg <- drone[d,1:Nloci]
      genotype <- offspring_genotype(Nloci, qg, dg)
      worker[w,1:(2*Nloci)] <- t(genotype)
    }   

    # Calculate pairwise relatedness
    gdata <- data.frame(names,worker)	# data frame with worker names and genotypes
    output <- coancestry(gdata, allele.freqs=allele_list, quellergt=1)
    meanRsim8[sim] <- mean(output$relatedness[[10]])	# calculate and save the mean pairwise R
    
  }

  # Calculate the p-value of the observed mean pairwise R for this colony relative to the simulated values
  Pval9[col] <- (rank(c(meanR,meanRsim8))[1])/(Nsim + 1)

  ########################################################################################################

}
as.data.frame(Pval3)
as.data.frame(Pval4)
as.data.frame(Pval5)
as.data.frame(Pval6)
as.data.frame(Pval7)
as.data.frame(Pval8)
as.data.frame(Pval9)


########################################################################################################
# Relatedness statistics for haplodiploid colonies (7 March 2018)

# This code generates random samples of daughter.queen genotypes assuming different degrees of 
# relatedness between the parents.
# Four unlinked microsatellite loci are simulated, using allele frequencies supplied
# by the user.


# Load packages and function files

library(related)
source("functions.r")

# Defined parameters

Nsim = 1000	# number of simulations used to estimate each p-value
Nloci = 4	# number of microsatellite loci
mf = 2		# number of male mates per queen

# Read in allele frequencies

freqs <- read.table("allele_freqs.txt", header=TRUE, sep="\t")
locus1 <- subset(freqs, locus == 1, select=c(allele, freq))
locus2 <- subset(freqs, locus == 2, select=c(allele, freq))
locus3 <- subset(freqs, locus == 3, select=c(allele, freq))
locus4 <- subset(freqs, locus == 4, select=c(allele, freq))
allele_list <- list(locus1,locus2,locus3,locus4)

# Read in colony relatedness statistics: all nestmate matrilines are siblings.

cdata <- read.table("colonies_inferred matrilines_2 motherqueens_somecolonies.txt", header=TRUE, sep="\t")

Ncolonies <- length(cdata$colony)	# number of colonies in data set


# Relatedness p-values

Pval1 = numeric(Ncolonies)
Pval2 = numeric(Ncolonies)

for (col in 1:Ncolonies) {

  Ndaughter.queens <- cdata$n[col]
  Nmother.queens <- cdata$q[col]
  Ndrones <- mf*Nmother.queens
  meanR <- cdata$R[col] 	# mean pairwise R for this colony
  names = c(1:Ndaughter.queens)
 

  ########################################################################################################
  # Scenario I: q unrelated mother queens each mated to mf unrelated fathers

  meanRsim1 <- numeric(Nsim)	# This array will contain the simulated mean pairwise R's

  for (sim in 1:Nsim)  {
    
    # Sample the unrelated queens and drones
 
    mother.queen <- matrix(0,Nmother.queens,2*Nloci)
    for (q in 1:Nmother.queens)  {
      genotype <- diploid_genotype(Nloci, allele_list)
      mother.queen[q,1:(2*Nloci)] <- t(genotype)
    }

    drone <- matrix(0,Ndrones,Nloci)
    for (d in 1:Ndrones)  {
      genotype <- haploid_genotype(Nloci, allele_list)
      drone[d,1:Nloci] <- t(genotype)
    }

    # Sample the offspring (daughter.queen) genotypes
    daughter.queen <- matrix(0,Ndaughter.queens,2*Nloci)
    for (w in 1:Ndaughter.queens) {
      q <- runif(1,1,Nmother.queens)  		# choose the mother
      d <- runif(1,mf*(q-1)+1,mf*q)  	# choose the father
      qg <- mother.queen[q,1:(2*Nloci)]
      dg <- drone[d,1:Nloci]
      genotype <- offspring_genotype(Nloci, qg, dg)
      daughter.queen[w,1:(2*Nloci)] <- t(genotype)
    }   

    # Calculate pairwise relatedness
    gdata <- data.frame(names,daughter.queen)	# data frame with daughter.queen names and genotypes
    output1 <- coancestry(gdata, allele.freqs=allele_list, quellergt=1)
    meanRsim1[sim] <- mean(output1$relatedness[[10]])	# calculate and save the mean pairwise R
  }

  # Calculate the p-value of the observed mean pairwise R for this colony relative to the simulated values
  Pval1[col] <- (rank(c(meanR,meanRsim1))[1])/(Nsim + 1)

  ########################################################################################################

  ########################################################################################################
  # Scenario II: q mother queens that all share a common mother (grand-queen) and mf unrelated fathers (grand-drones)
  # (i.e. all daughter queens are first cousins)

  meanRsim2 <- numeric(Nsim)	# This array will contain the simulated mean pairwise R's

  for (sim in 1:Nsim)  {
    
    # Sample the grand-queen and her male mates

    grandqueen <- diploid_genotype(Nloci, allele_list)
    granddrone <- matrix(0,mf,Nloci)
    for (d in 1:mf)  {
      genotype <- haploid_genotype(Nloci, allele_list)
      granddrone[d,1:Nloci] <- t(genotype)
    }

    # Sample the mother queens and their unrelated male mates
 
    mother.queen <- matrix(0,Nmother.queens,2*Nloci)
    for (q in 1:Nmother.queens)  {
      qg <- grandqueen
      d <- runif(1,1,mf)  	# choose the father (granddrone)
      dg <- granddrone[d,1:Nloci]
      genotype <- offspring_genotype(Nloci, qg, dg)
      mother.queen[q,1:(2*Nloci)] <- t(genotype)
    }

    drone <- matrix(0,Ndrones,Nloci)
    for (d in 1:Ndrones)  {
      genotype <- haploid_genotype(Nloci, allele_list)
      drone[d,1:Nloci] <- t(genotype)
    }

    # Sample the offspring (daughter queen) genotypes
    daughter.queen <- matrix(0,Ndaughter.queens,2*Nloci)
    for (w in 1:Ndaughter.queens) {
      q <- runif(1,1,Nmother.queens)  		# choose the motherqueen 
      d <- runif(1,mf*(q-1)+1,mf*q)  	# choose the father
      qg <- mother.queen[q,1:(2*Nloci)]
      dg <- drone[d,1:Nloci]
      genotype <- offspring_genotype(Nloci, qg, dg)
      daughter.queen[w,1:(2*Nloci)] <- t(genotype)
    }   

    # Calculate pairwise relatedness
    gdata <- data.frame(names,daughter.queen)	# data frame with daughter.queen names and genotypes
    output2 <- coancestry(gdata, allele.freqs=allele_list, quellergt=1)
    meanRsim2[sim] <- mean(output2$relatedness[[10]])	# calculate and save the mean pairwise R
  }

  # Calculate the p-value of the observed mean pairwise R for this colony relative to the simulated values
  Pval2[col] <- (rank(c(meanR,meanRsim2))[1])/(Nsim + 1)
 ########################################################################################################

 
}

as.data.frame(Pval1)
as.data.frame(Pval2)



########################################################################################################
# Relatedness statistics for haplodiploid colonies (7 March 2018)

# This code generates random samples of worker genotypes assuming different degrees of 
# relatedness between the parents.
# Four unlinked microsatellite loci are simulated, using allele frequencies supplied
# by the user.


# Load packages and function files

library(related)
source("functions.r")


# Defined parameters

Nsim = 1000	# number of simulations used to estimate each p-value
Nloci = 4	# number of microsatellite loci
mf = 2		# number of male mates per queen
grandqueen2<-matrix(ncol=8,nrow=2)
grandqueen3<-matrix(ncol=8,nrow=3)
grandqueen8<-matrix(ncol=8,nrow=8)



# Read in allele frequencies

freqs <- read.table("allele_freqs.txt", header=TRUE, sep="\t")
locus1 <- subset(freqs, locus == 1, select=c(allele, freq))
locus2 <- subset(freqs, locus == 2, select=c(allele, freq))
locus3 <- subset(freqs, locus == 3, select=c(allele, freq))
locus4 <- subset(freqs, locus == 4, select=c(allele, freq))
allele_list <- list(locus1,locus2,locus3,locus4)

# Read in colony relatedness statistics

cdata <- read.table("colonies_inferred matrilines_1 motherqueen.txt", header=TRUE, sep="\t")

Ncolonies <- length(cdata$colony)	# number of colonies in data set


# Relatedness p-values

Pval3 = numeric(Ncolonies)
Pval4 = numeric(Ncolonies)

for (col in 1:Ncolonies) {

  Ndaugther.queens <- cdata$n[col]
  Nmother.queens <- cdata$q[col]
  Ndrones <- mf*Nmother.queens
  meanR <- cdata$R[col] 	# mean pairwise R for this colony
  names = c(1:Ndaugther.queens)
 

 ########################################################################################################
  # Scenario III: q mother queens that share 2 common mothers and mf unrelated fathers 

  meanRsim3 <- numeric(Nsim)	# This array will contain the simulated mean pairwise R's

  for (sim in 1:Nsim)  {
    

    # Sample the two grand-queens and their male mates

    grandqueen2[1,1:(2*Nloci)] <- diploid_genotype(Nloci, allele_list)
    grandqueen2[2,1:(2*Nloci)] <- diploid_genotype(Nloci, allele_list)
    granddrone <- matrix(0,2*mf,Nloci)
    for (d in 1:(2*mf))  {
      genotype <- haploid_genotype(Nloci, allele_list)
      granddrone[d,1:Nloci] <- t(genotype)
    }

    # Sample the queens and their unrelated male mates
 
    queen <- matrix(0,Nmother.queens,2*Nloci)
    for (q in 1:Nmother.queens)  {
      g <- runif(1,1,2)		# choose the mother (grandquuen)
      qg <- grandqueen2[g,1:(2*Nloci)]
      d <- runif(1,1,mf)  	# choose the father (granddrone)
      dg <- granddrone[(g-1)*mf+d,1:Nloci]
      genotype <- offspring_genotype(Nloci, qg, dg)
      queen[q,1:(2*Nloci)] <- t(genotype)
    }

    drone <- matrix(0,Ndrones,Nloci)
    for (d in 1:Ndrones)  {
      genotype <- haploid_genotype(Nloci, allele_list)
      drone[d,1:Nloci] <- t(genotype)
    }

    # Sample the offspring (worker) genotypes
    worker <- matrix(0,Ndaugther.queens,2*Nloci)
    for (w in 1:Ndaugther.queens) {
      q <- runif(1,1,Nmother.queens)  		# choose the mother
      d <- runif(1,mf*(q-1)+1,mf*q)  	# choose the father
      qg <- queen[q,1:(2*Nloci)]
      dg <- drone[d,1:Nloci]
      genotype <- offspring_genotype(Nloci, qg, dg)
      worker[w,1:(2*Nloci)] <- t(genotype)
    }   

    # Calculate pairwise relatedness
    gdata <- data.frame(names,worker)	# data frame with worker names and genotypes
    output <- coancestry(gdata, allele.freqs=allele_list, quellergt=1)
    meanRsim3[sim] <- mean(output$relatedness[[10]])	# calculate and save the mean pairwise R
    
  }

  # Calculate the p-value of the observed mean pairwise R for this colony relative to the simulated values
  Pval3[col] <- (rank(c(meanR,meanRsim3))[1])/(Nsim + 1)

  ########################################################################################################


  ########################################################################################################
  # Scenario 4: q queens that share 3 common mothers and mf unrelated fathers 

  meanRsim4 <- numeric(Nsim)	# This array will contain the simulated mean pairwise R's

  for (sim in 1:Nsim)  {
    

    # Sample the two grand-queens and their male mates

    grandqueen3[1,1:(2*Nloci)] <- diploid_genotype(Nloci, allele_list)
    grandqueen3[2,1:(2*Nloci)] <- diploid_genotype(Nloci, allele_list)
    grandqueen3[3,1:(2*Nloci)] <- diploid_genotype(Nloci, allele_list)
    granddrone <- matrix(0,3*mf,Nloci)
    for (d in 1:(3*mf))  {
      genotype <- haploid_genotype(Nloci, allele_list)
      granddrone[d,1:Nloci] <- t(genotype)
    }

    # Sample the queens and their unrelated male mates
 
    queen <- matrix(0,Nmother.queens,2*Nloci)
    for (q in 1:Nmother.queens)  {
      g <- runif(1,1,3)		# choose the mother (grandquuen)
      qg <- grandqueen3[g,1:(2*Nloci)]
      d <- runif(1,1,mf)  	# choose the father (granddrone)
      dg <- granddrone[(g-1)*mf+d,1:Nloci]
      genotype <- offspring_genotype(Nloci, qg, dg)
      queen[q,1:(2*Nloci)] <- t(genotype)
    }

    drone <- matrix(0,Ndrones,Nloci)
    for (d in 1:Ndrones)  {
      genotype <- haploid_genotype(Nloci, allele_list)
      drone[d,1:Nloci] <- t(genotype)
    }

    # Sample the offspring (worker) genotypes
    worker <- matrix(0,Ndaugther.queens,2*Nloci)
    for (w in 1:Ndaugther.queens) {
      q <- runif(1,1,Nmother.queens)  		# choose the mother
      d <- runif(1,mf*(q-1)+1,mf*q)  	# choose the father
      qg <- queen[q,1:(2*Nloci)]
      dg <- drone[d,1:Nloci]
      genotype <- offspring_genotype(Nloci, qg, dg)
      worker[w,1:(2*Nloci)] <- t(genotype)
    }   

    # Calculate pairwise relatedness
    gdata <- data.frame(names,worker)	# data frame with worker names and genotypes
    output <- coancestry(gdata, allele.freqs=allele_list, quellergt=1)
    meanRsim4[sim] <- mean(output$relatedness[[10]])	# calculate and save the mean pairwise R
    
  }

  # Calculate the p-value of the observed mean pairwise R for this colony relative to the simulated values
  Pval4[col] <- (rank(c(meanR,meanRsim4))[1])/(Nsim + 1)

  ########################################################################################################

}
as.data.frame(Pval3)
as.data.frame(Pval4)


#end#