######################################################################################################### ################################## Simulation script for manuscript: #################################### ####### Strategies to improve selection compared to selection based on estimated breeding values ###### ######################################################################################################### ## Using args to submit different strategies within the same shell script args <- commandArgs(TRUE) scenario <- as.numeric(args[2]) seed <- as.numeric(args[1]) # some scenarios make use of currently WUR-internal software for estimation of # Mendelian sampling variance. Formulas to derive MSV are given in: # Niehoff et al. (https://gsejournal.biomedcentral.com/articles/10.1186/s12711-024-00899-2#additional-information) # less efficient code is given in: # Niehoff et al. (https://academic.oup.com/g3journal/article/14/11/jkae205/7743298?login=true#492554989) # directory to write results to title = paste0("uniqueness_v3/sc", scenario, "seed", seed, "_potential.RData") library(MoBPS) # Basic population parameters n_additive = 1000 # Number of additive QTLs heritability = 0.3 # heritability pop_size = c(500, 500) # number of males / females per cycle sel_size = c(40, 100) # number of selected males / females per cycle nsnp = 25000 # number of SNPs chr = 10 # number of chromosomes chr_length = 2.5 # length of each chromosome in Morgan burnin = 10 cycles = 50 weight_unique = 0.05 display.progress = FALSE ## Generation of the baseline population set.seed(seed) population = creating.diploid(nsnp = nsnp, nindi = sel_size, chr.nr = chr, chromosome.length = chr_length, n.additive = n_additive, share.genotyped = 1) alpha_real = numeric(nsnp) alpha_real[population$info$real.bv.add[[1]][,6]] = (population$info$real.bv.add[[1]][,5] - population$info$real.bv.add[[1]][,3])/2 # Burn-in phase population = breeding.diploid(population, heritability = heritability) for(index in 1:burnin){ population = breeding.diploid(population, phenotyping.gen = get.ngen(population)) population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "pheno", share.genotyped = 1, display.progress = display.progress) } # Standardization of traits in the first generation (after burn-in) population = bv.standardization(population, mean.target = 100, var.target = 1, gen = get.ngen(population), adapt.pheno = TRUE) population = breeding.diploid(population, heritability = heritability) for(index in 1:cycles){ current_gen = get.ngen(population) population = breeding.diploid(population, phenotyping.gen = current_gen) bve.gen = bve.gen1 = current_gen if(scenario %in% 335:338){ bve.gen = (current_gen-5):current_gen bve.gen1 = current_gen } population = breeding.diploid(population, bve = TRUE, bve.gen = bve.gen, estimate.u = TRUE) alpha_hat = population$info$u_hat[[length(population$info$u_hat)]][,1] # estimated SNP effects if(sum(is.na(alpha_hat))>0){ alpha_hat[is.na(alpha_hat)] = 0 } geno = get.geno(population = population, gen = current_gen) # genotypes p = rowMeans(geno)/2 # frequency of "good" allele freq_alpha = p freq_alpha[alpha_real==0] = 0 # these can be ignored freq_alpha[alpha_real<0] = 1 -freq_alpha[alpha_real<0] # these can be ignored # frequency of estimated "good" allele freq_alphahat = p freq_alphahat[alpha_hat==0] = 0 # these can be ignored freq_alphahat[alpha_hat<0] = 1 -freq_alphahat[alpha_hat<0] # these can be ignored ebv = get.bve(population, gen = current_gen)[1,] # estimated breeding value pheno = get.pheno(population, gen = current_gen)[1,] # phenotype bv = get.bv(population, gen = current_gen)[1,] # true underlying genomic value # this script was limited to those scenarios presented in the manuscript # scenario numbers were not changes (explaining why there is no scenario 2, 4, 6, 7, etc.) # order of scenarios corresponds to the other of results given in Supplementary Tables S1-S6 if( scenario == 1){ # Selection based on genomic breeding value population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, display.progress = display.progress) } else if(scenario == 3){ sel_size_tmp = sel_size # selection on the female side is done without OGC, on male male size selection.size = population size # to provide optiSel with all potential sires # calculation on how to set ub.sKin can be tricky has this corresponds to the increase of inbreeding # that is caused by the male side of the breeding program sel_size_tmp[1] = pop_size[1] # Selection based on genomic breeding value population = breeding.diploid(population, selection.size = sel_size_tmp, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, ogc = TRUE, ogc.target = "max.BV", ogc.uniform = "female", ogc.ub.sKin.increase = 0.005, display.progress = display.progress) } else if(scenario == 5){ activ = alpha_hat!= 0 weighting = sqrt(1 / freq_alphahat) weighting[weighting==Inf] = 0 unique = as.numeric((alpha_hat[activ] * weighting[activ]) %*% geno[activ,]) new_ebv = (unique - mean(unique))/sd(unique) population = insert.bve(population, bves = cbind(names(ebv), new_ebv)) population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, display.progress = display.progress) } else if( scenario == 10){ cand = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", export.selected = TRUE ) A = kinship.exp(population, gen = get.ngen(population)) rel_top = colMeans(A[cand[[1]][,3],]) + colMeans(A[cand[[2]][,3] + pop_size[1],]) unique = -rel_top # new_ebv = weight_unique * (unique - mean(unique))/sd(unique) + (1-weight_unique)* (ebv - mean(ebv))/sd(ebv) population = insert.bve(population, bves = cbind(names(ebv), new_ebv)) population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, display.progress = display.progress) } else if( scenario == 14){ weight_unique = 0.2 cand = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", export.selected = TRUE ) A = kinship.exp(population, gen = get.ngen(population)) rel_top = colMeans(A[cand[[1]][,3],]) + colMeans(A[cand[[2]][,3] + pop_size[1],]) unique = -rel_top # new_ebv = weight_unique * (unique - mean(unique))/sd(unique) + (1-weight_unique)* (ebv - mean(ebv))/sd(ebv) population = insert.bve(population, bves = cbind(names(ebv), new_ebv)) population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, display.progress = display.progress) } else if( scenario == 36){ cand = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", export.selected = TRUE ) A = kinship.exp(population, gen = get.ngen(population)) rel_top = colMeans(A) unique = -rel_top # new_ebv = weight_unique * (unique - mean(unique))/sd(unique) + (1-weight_unique)* (ebv - mean(ebv))/sd(ebv) population = insert.bve(population, bves = cbind(names(ebv), new_ebv)) population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, display.progress = display.progress) } else if( scenario == 37){ weight_unique = 0.2 cand = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", export.selected = TRUE ) A = kinship.exp(population, gen = get.ngen(population)) rel_top = colMeans(A) unique = -rel_top # new_ebv = weight_unique * (unique - mean(unique))/sd(unique) + (1-weight_unique)* (ebv - mean(ebv))/sd(ebv) population = insert.bve(population, bves = cbind(names(ebv), new_ebv)) population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, display.progress = display.progress) } else if( scenario == 53){ weight_unique = 0.025 cand = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", export.selected = TRUE ) A = kinship.exp(population, gen = get.ngen(population)) diag(A) = (diag(A)-0.5)*2 rel_top = colMeans(A[cand[[1]][,3],]) + colMeans(A[cand[[2]][,3] + pop_size[1],]) unique = -rel_top # new_ebv = weight_unique * (unique - mean(unique))/sd(unique) + (1-weight_unique)* (ebv - mean(ebv))/sd(ebv) population = insert.bve(population, bves = cbind(names(ebv), new_ebv)) population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, display.progress = display.progress) } else if( scenario == 54){ weight_unique = 0.075 cand = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", export.selected = TRUE ) A = kinship.exp(population, gen = get.ngen(population)) diag(A) = (diag(A)-0.5)*2 rel_top = colMeans(A[cand[[1]][,3],]) + colMeans(A[cand[[2]][,3] + pop_size[1],]) unique = -rel_top # new_ebv = weight_unique * (unique - mean(unique))/sd(unique) + (1-weight_unique)* (ebv - mean(ebv))/sd(ebv) population = insert.bve(population, bves = cbind(names(ebv), new_ebv)) population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, display.progress = display.progress) } else if( scenario == 55){ weight_unique = 0.1 cand = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", export.selected = TRUE ) A = kinship.exp(population, gen = get.ngen(population)) diag(A) = (diag(A)-0.5)*2 rel_top = colMeans(A[cand[[1]][,3],]) + colMeans(A[cand[[2]][,3] + pop_size[1],]) unique = -rel_top # new_ebv = weight_unique * (unique - mean(unique))/sd(unique) + (1-weight_unique)* (ebv - mean(ebv))/sd(ebv) population = insert.bve(population, bves = cbind(names(ebv), new_ebv)) population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, display.progress = display.progress) } else if( scenario == 56){ weight_unique = 0.125 cand = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", export.selected = TRUE ) A = kinship.exp(population, gen = get.ngen(population)) diag(A) = (diag(A)-0.5)*2 rel_top = colMeans(A[cand[[1]][,3],]) + colMeans(A[cand[[2]][,3] + pop_size[1],]) unique = -rel_top # new_ebv = weight_unique * (unique - mean(unique))/sd(unique) + (1-weight_unique)* (ebv - mean(ebv))/sd(ebv) population = insert.bve(population, bves = cbind(names(ebv), new_ebv)) population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, display.progress = display.progress) }else if( scenario == 57){ weight_unique = 0.15 cand = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", export.selected = TRUE ) A = kinship.exp(population, gen = get.ngen(population)) diag(A) = (diag(A)-0.5)*2 rel_top = colMeans(A[cand[[1]][,3],]) + colMeans(A[cand[[2]][,3] + pop_size[1],]) unique = -rel_top # new_ebv = weight_unique * (unique - mean(unique))/sd(unique) + (1-weight_unique)* (ebv - mean(ebv))/sd(ebv) population = insert.bve(population, bves = cbind(names(ebv), new_ebv)) population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, display.progress = display.progress) } else if( scenario == 58){ weight_unique = 0.175 cand = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", export.selected = TRUE ) A = kinship.exp(population, gen = get.ngen(population)) diag(A) = (diag(A)-0.5)*2 rel_top = colMeans(A[cand[[1]][,3],]) + colMeans(A[cand[[2]][,3] + pop_size[1],]) unique = -rel_top # new_ebv = weight_unique * (unique - mean(unique))/sd(unique) + (1-weight_unique)* (ebv - mean(ebv))/sd(ebv) population = insert.bve(population, bves = cbind(names(ebv), new_ebv)) population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, display.progress = display.progress) } else if( scenario == 59){ weight_unique = 0.25 cand = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", export.selected = TRUE ) A = kinship.exp(population, gen = get.ngen(population)) diag(A) = (diag(A)-0.5)*2 rel_top = colMeans(A[cand[[1]][,3],]) + colMeans(A[cand[[2]][,3] + pop_size[1],]) unique = -rel_top # new_ebv = weight_unique * (unique - mean(unique))/sd(unique) + (1-weight_unique)* (ebv - mean(ebv))/sd(ebv) population = insert.bve(population, bves = cbind(names(ebv), new_ebv)) population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, display.progress = display.progress) }else if( scenario == 60){ weight_unique = 0.3 cand = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", export.selected = TRUE ) A = kinship.exp(population, gen = get.ngen(population)) diag(A) = (diag(A)-0.5)*2 rel_top = colMeans(A[cand[[1]][,3],]) + colMeans(A[cand[[2]][,3] + pop_size[1],]) unique = -rel_top # new_ebv = weight_unique * (unique - mean(unique))/sd(unique) + (1-weight_unique)* (ebv - mean(ebv))/sd(ebv) population = insert.bve(population, bves = cbind(names(ebv), new_ebv)) population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, display.progress = display.progress) } else if( scenario == 61){ weight_unique = 0.4 cand = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", export.selected = TRUE ) A = kinship.exp(population, gen = get.ngen(population)) diag(A) = (diag(A)-0.5)*2 rel_top = colMeans(A[cand[[1]][,3],]) + colMeans(A[cand[[2]][,3] + pop_size[1],]) unique = -rel_top # new_ebv = weight_unique * (unique - mean(unique))/sd(unique) + (1-weight_unique)* (ebv - mean(ebv))/sd(ebv) population = insert.bve(population, bves = cbind(names(ebv), new_ebv)) population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, display.progress = display.progress) } else if( scenario == 62){ weight_unique = 0.5 cand = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", export.selected = TRUE ) A = kinship.exp(population, gen = get.ngen(population)) diag(A) = (diag(A)-0.5)*2 rel_top = colMeans(A[cand[[1]][,3],]) + colMeans(A[cand[[2]][,3] + pop_size[1],]) unique = -rel_top # new_ebv = weight_unique * (unique - mean(unique))/sd(unique) + (1-weight_unique)* (ebv - mean(ebv))/sd(ebv) population = insert.bve(population, bves = cbind(names(ebv), new_ebv)) population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, display.progress = display.progress) } else if( scenario == 95){ # Selection based on genomic breeding value population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, avoid.mating.fullsib = TRUE, display.progress = display.progress) } else if( scenario == 96){ # Selection based on genomic breeding value population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, avoid.mating.halfsib = TRUE, display.progress = display.progress) } else if( scenario == 97){ # Selection based on genomic breeding value population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, avoid.mating.inb.quantile = 0.9, display.progress = display.progress) } else if( scenario == 98){ # Selection based on genomic breeding value population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, avoid.mating.inb.quantile = 0.5, display.progress = display.progress) } else if( scenario == 99){ # Selection based on genomic breeding value population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, avoid.mating.inb.quantile = 0.25, display.progress = display.progress) } else if( scenario == 100){ # Selection based on genomic breeding value population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, avoid.mating.inb.quantile = 0.1, display.progress = display.progress) } else if( scenario == 118){ # Selection based on genomic breeding value population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, avoid.mating.kinship.quantile = 0.9, avoid.mating.kinship.gen = get.ngen(population), display.progress = display.progress) } else if( scenario == 119){ # Selection based on genomic breeding value population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, avoid.mating.kinship.quantile = 0.5, avoid.mating.kinship.gen = get.ngen(population), display.progress = display.progress) } else if( scenario == 120){ # Selection based on genomic breeding value population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, avoid.mating.kinship.quantile = 0.25, avoid.mating.kinship.gen = get.ngen(population), display.progress = display.progress) } else if( scenario == 121){ # Selection based on genomic breeding value population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, avoid.mating.kinship.quantile = 0.1, avoid.mating.kinship.gen = get.ngen(population), display.progress = display.progress) } else if( scenario == 125){ # Selection based on genomic breeding value population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, avoid.mating.inb.quantile = 0.8, display.progress = display.progress) } else if( scenario == 126){ # Selection based on genomic breeding value population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, avoid.mating.inb.quantile = 0.7, display.progress = display.progress) } else if( scenario == 127){ # Selection based on genomic breeding value population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, avoid.mating.inb.quantile = 0.6, display.progress = display.progress) } else if( scenario == 130){ # Selection based on genomic breeding value population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, avoid.mating.inb.quantile = 0.4, display.progress = display.progress) } else if( scenario == 131){ # Selection based on genomic breeding value population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, avoid.mating.inb.quantile = 0.3, display.progress = display.progress) } else if( scenario == 132){ # Selection based on genomic breeding value population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, avoid.mating.inb.quantile = 0.2, display.progress = display.progress) } else if( scenario == 133){ # Selection based on genomic breeding value population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, avoid.mating.kinship.quantile = 0.8, avoid.mating.kinship.gen = get.ngen(population), display.progress = display.progress) } else if( scenario == 134){ # Selection based on genomic breeding value population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, avoid.mating.kinship.quantile = 0.7, avoid.mating.kinship.gen = get.ngen(population), display.progress = display.progress) } else if( scenario == 135){ # Selection based on genomic breeding value population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, avoid.mating.kinship.quantile = 0.6, avoid.mating.kinship.gen = get.ngen(population), display.progress = display.progress) }else if( scenario == 136){ # Selection based on genomic breeding value population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, avoid.mating.kinship.quantile = 0.4, avoid.mating.kinship.gen = get.ngen(population), display.progress = display.progress) }else if( scenario == 137){ # Selection based on genomic breeding value population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, avoid.mating.kinship.quantile = 0.3, avoid.mating.kinship.gen = get.ngen(population), display.progress = display.progress) }else if( scenario == 138){ # Selection based on genomic breeding value population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, avoid.mating.kinship.quantile = 0.2, avoid.mating.kinship.gen = get.ngen(population), display.progress = display.progress) } else if(scenario == 187){ activ = alpha_hat!= 0 weighting = (1 / freq_alphahat)^(1/2) weighting[weighting==Inf] = 0 unique = as.numeric((alpha_hat[activ] * weighting[activ]) %*% geno[activ,]) new_ebv = (unique - mean(unique))/sd(unique) population = insert.bve(population, bves = cbind(names(ebv), new_ebv)) population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, display.progress = display.progress) } else if(scenario == 188){ activ = alpha_hat!= 0 weighting = (1 / freq_alphahat)^(1/2) weighting[weighting==Inf] = 0 weighting[weighting > 5] = 5 unique = as.numeric((alpha_hat[activ] * weighting[activ]) %*% geno[activ,]) new_ebv = (unique - mean(unique))/sd(unique) population = insert.bve(population, bves = cbind(names(ebv), new_ebv)) population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, display.progress = display.progress) } else if(scenario == 189){ activ = alpha_hat!= 0 weighting = (1 / freq_alphahat)^(1/3) weighting[weighting==Inf] = 0 unique = as.numeric((alpha_hat[activ] * weighting[activ]) %*% geno[activ,]) new_ebv = (unique - mean(unique))/sd(unique) population = insert.bve(population, bves = cbind(names(ebv), new_ebv)) population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, display.progress = display.progress) } else if(scenario == 190){ activ = alpha_hat!= 0 weighting = (1 / freq_alphahat)^(1/4) weighting[weighting==Inf] = 0 unique = as.numeric((alpha_hat[activ] * weighting[activ]) %*% geno[activ,]) new_ebv = (unique - mean(unique))/sd(unique) population = insert.bve(population, bves = cbind(names(ebv), new_ebv)) population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, display.progress = display.progress) } else if(scenario == 191){ activ = alpha_hat!= 0 weighting = 0.8 + 0.2 * (1 / freq_alphahat)^(1/2) weighting[weighting==Inf] = 0 unique = as.numeric((alpha_hat[activ] * weighting[activ]) %*% geno[activ,]) new_ebv = (unique - mean(unique))/sd(unique) population = insert.bve(population, bves = cbind(names(ebv), new_ebv)) population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, display.progress = display.progress) } else if(scenario == 192){ activ = alpha_hat!= 0 weighting = 0.6 + 0.4 * (1 / freq_alphahat)^(1/2) weighting[weighting==Inf] = 0 unique = as.numeric((alpha_hat[activ] * weighting[activ]) %*% geno[activ,]) new_ebv = (unique - mean(unique))/sd(unique) population = insert.bve(population, bves = cbind(names(ebv), new_ebv)) population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, display.progress = display.progress) } else if(scenario == 193){ activ = alpha_hat!= 0 weighting = (pi/2 - asin(sqrt(freq_alphahat)))/(sqrt(freq_alphahat * (1-freq_alphahat))) weighting[weighting==Inf] = 0 unique = as.numeric((alpha_hat[activ] * weighting[activ]) %*% geno[activ,]) new_ebv = (unique - mean(unique))/sd(unique) population = insert.bve(population, bves = cbind(names(ebv), new_ebv)) population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, display.progress = display.progress) } else if( scenario == 231){ work_dir = paste0("mixblup_", scenario, "_", seed) if(index == 1){ dir.create(work_dir) system(paste0("cp /home/WUR/pook001/SysDir.inp ", work_dir, "/SysDir.inp")) setwd(work_dir) system(paste0("cp /home/WUR/pook001/msv msv")) system("chmod +x msv") } get.plink(population, gen =bve.gen, path = "paternal_plink", fam.id = TRUE, type = 1) get.plink(population, gen =bve.gen, path = "maternal_plink", fam.id = TRUE, type = 2) write.table(file = "EBV.txt", cbind(get.id(population, gen = bve.gen), ebv), row.names = FALSE, col.names = FALSE, quote = FALSE) write.table(file = "sol.dat", cbind(1,1,1,1:length(alpha_hat), alpha_hat), row.names = FALSE, col.names = FALSE, quote = FALSE) write.table(file = "all_freq", cbind(1:length(alpha_hat), 0), row.names = FALSE, col.names = FALSE, quote = FALSE) system("./msv --matgeno maternal_plink.bed --patgeno paternal_plink.bed --freq all_freq --ebv EBV.txt --sol sol.dat --snpeff 1 --nrgen 2 --selprop 0.1 --bp2M 1e-08") a = read.table("EBV.txt") b = read.table("genetic_MSVs_1gen.txt") c = read.table("Index_5_6.txt") #Piter Index 5/6 d = read.table("ExpBVSelGrOff_per_individual.txt") # Tobias Index new_ebv = d[,2] insert.bve = cbind(names(ebv), new_ebv) population = insert.bve(population, bves = insert.bve ) acc2 = rbind(acc2, t(analyze.bv(population, gen = bve.gen)[[1]])) pop1 = breeding.diploid(population, bve = TRUE, bve.gen = bve.gen) acc1 = rbind(acc1, t(analyze.bv(pop1, gen = bve.gen)[[1]])) print(colMeans(acc1)) print(colMeans(acc2)) # Selection based on genomic breeding value population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, display.progress = display.progress) } else if( scenario == 233){ work_dir = paste0("mixblup_", scenario, "_", seed) if(index == 1){ dir.create(work_dir) system(paste0("cp /home/WUR/pook001/SysDir.inp ", work_dir, "/SysDir.inp")) setwd(work_dir) system(paste0("cp /home/WUR/pook001/msv msv")) system("chmod +x msv") } get.plink(population, gen =bve.gen, path = "paternal_plink", fam.id = TRUE, type = 1) get.plink(population, gen =bve.gen, path = "maternal_plink", fam.id = TRUE, type = 2) write.table(file = "EBV.txt", cbind(get.id(population, gen = bve.gen), ebv), row.names = FALSE, col.names = FALSE, quote = FALSE) write.table(file = "sol.dat", cbind(1,1,1,1:length(alpha_hat), alpha_hat), row.names = FALSE, col.names = FALSE, quote = FALSE) write.table(file = "all_freq", cbind(1:length(alpha_hat), 0), row.names = FALSE, col.names = FALSE, quote = FALSE) system("./msv --matgeno maternal_plink.bed --patgeno paternal_plink.bed --freq all_freq --ebv EBV.txt --sol sol.dat --snpeff 1 --nrgen 2 --selprop 0.1 --bp2M 1e-08") a = read.table("EBV.txt") b = read.table("genetic_MSVs_1gen.txt") c = read.table("Index_5_6.txt") #Piter Index 5/6 d = read.table("ExpBVSelGrOff_per_individual.txt") # Tobias Index new_ebv = c[,2] insert.bve = cbind(names(ebv), new_ebv) population = insert.bve(population, bves = insert.bve ) acc2 = rbind(acc2, t(analyze.bv(population, gen = bve.gen)[[1]])) pop1 = breeding.diploid(population, bve = TRUE, bve.gen = bve.gen) acc1 = rbind(acc1, t(analyze.bv(pop1, gen = bve.gen)[[1]])) print(colMeans(acc1)) print(colMeans(acc2)) # Selection based on genomic breeding value population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, display.progress = display.progress) } else if( scenario == 234){ work_dir = paste0("mixblup_", scenario, "_", seed) if(index == 1){ dir.create(work_dir) system(paste0("cp /home/WUR/pook001/SysDir.inp ", work_dir, "/SysDir.inp")) setwd(work_dir) system(paste0("cp /home/WUR/pook001/msv msv")) system("chmod +x msv") } get.plink(population, gen =bve.gen, path = "paternal_plink", fam.id = TRUE, type = 1) get.plink(population, gen =bve.gen, path = "maternal_plink", fam.id = TRUE, type = 2) write.table(file = "EBV.txt", cbind(get.id(population, gen = bve.gen), ebv), row.names = FALSE, col.names = FALSE, quote = FALSE) write.table(file = "sol.dat", cbind(1,1,1,1:length(alpha_hat), alpha_hat), row.names = FALSE, col.names = FALSE, quote = FALSE) write.table(file = "all_freq", cbind(1:length(alpha_hat), 0), row.names = FALSE, col.names = FALSE, quote = FALSE) system("./msv --matgeno maternal_plink.bed --patgeno paternal_plink.bed --freq all_freq --ebv EBV.txt --sol sol.dat --snpeff 1 --nrgen 2 --selprop 0.1 --bp2M 1e-08") a = read.table("EBV.txt") b = read.table("genetic_MSVs_1gen.txt") c = read.table("Index_5_6.txt") #Piter Index 5/6 d = read.table("ExpBVSelGrOff_per_individual.txt") # Tobias Index new_ebv = c[,3] insert.bve = cbind(names(ebv), new_ebv) population = insert.bve(population, bves = insert.bve ) acc2 = rbind(acc2, t(analyze.bv(population, gen = bve.gen)[[1]])) pop1 = breeding.diploid(population, bve = TRUE, bve.gen = bve.gen) acc1 = rbind(acc1, t(analyze.bv(pop1, gen = bve.gen)[[1]])) print(colMeans(acc1)) print(colMeans(acc2)) # Selection based on genomic breeding value population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, display.progress = display.progress) } else if( scenario == 237){ targetdegree = 5 temp_geno <- geno pi1 <- rowMeans(temp_geno)/2 temp_kin <- crossprod(temp_geno- 2*pi1) / sum(pi1*(1-pi1)) / 2 selection_index <- as.numeric(ebv) diag_kin <- diag(temp_kin) kin_mat <- temp_kin - min(temp_kin) kin_mat <- kin_mat/max(as.vector(kin_mat))*2 kin_mat_tri <- kin_mat[upper.tri(kin_mat)] kin_mat <- t(kin_mat) kin_mat[upper.tri(kin_mat)] <- kin_mat_tri dir.create(paste0("alphamaterun/unique_", targetdegree, "_", scenario,"_", seed)) write.table(kin_mat, sprintf("./alphamaterun/unique_%i_%i_%i/Nrm.txt", targetdegree,scenario,seed), sep="\t", row.names=TRUE, col.names=FALSE) criterion <- data.frame(Genotype = names(ebv), Value = as.vector(ebv)) write.table(criterion, sprintf("./alphamaterun/unique_%i_%i_%i/Criterion.txt", targetdegree,scenario,seed), sep="\t", row.names=FALSE, quote=FALSE, col.names=FALSE) gender_file = cbind(names(ebv), c(rep(1, pop_size[1]), rep(2, pop_size[2]))) write.table(gender_file, sprintf("./alphamaterun/unique_%i_%i_%i/Gender.txt", targetdegree,scenario,seed), sep="\t", row.names=FALSE, col.names=FALSE, quote = FALSE) spec_text <- sprintf("Seed , %i\nNrmMatrixFile , Nrm.txt\nGenderFile , Gender.txt\nSelCriterionFile , Criterion.txt\nNumberOfMatings , %i\nNumberOfMaleParents , %i\nNumberOfFemaleParents , %i\nTargetDegree , %i\nEqualizeMaleContributions , Yes\nEqualizeFemaleContributions , Yes\nPreselectMales , Yes\nPreselectMalePercentage , 100\nPreselectFemales , Yes\nPreselectFemalePercentage , 100\nEvolAlgNumberOfIterations , 1000\nEvolAlgNumberOfIterationsPrint , 5\nEvolAlgNumberOfIterationsStop , 50\nStop", seed, sum(pop_size), sel_size[1], sel_size[2],targetdegree) write(x = spec_text, file=sprintf("./alphamaterun/unique_%i_%i_%i/AlphaMateSpec.txt", targetdegree, scenario,seed)) sim_path = getwd() setwd(paste0("alphamaterun/unique_", targetdegree, "_", scenario,"_", seed)) suppressWarnings(file.remove("MatingPlanModeOptTarget1.txt")) if(seed<3){ system(sprintf(paste0( sim_path, "/AlphaMate/binaries/AlphaMate_Unix"))) } else{ system(sprintf(paste0( sim_path, "/AlphaMate_0.2.0/bin/AlphaMate_Linux_0.2.02"))) } setwd(sim_path) matings <- read.table(sprintf("./alphamaterun/unique_%i_%i_%i/MatingPlanModeOptTarget1.txt", targetdegree, scenario, seed), sep="", header=TRUE) males = match(as.character(matings$Parent1), names(ebv)[1:pop_size[1]]) females = match(as.character(matings$Parent2), names(ebv)[-(1:pop_size[1])]) sex = rep(0, sum(pop_size)) sex[sample(sum(pop_size), pop_size[2])] = 1 fixed_breeding = cbind(get.ngen(population), 1, males, get.ngen(population),2,females, sex) # Selection based on genomic breeding value population = breeding.diploid(population, fixed.breeding = fixed_breeding, share.genotyped = 1, display.progress = display.progress) }else if( scenario == 238){ targetdegree = 15 temp_geno <- geno pi1 <- rowMeans(temp_geno)/2 temp_kin <- crossprod(temp_geno- 2*pi1) / sum(pi1*(1-pi1)) / 2 selection_index <- as.numeric(ebv) diag_kin <- diag(temp_kin) kin_mat <- temp_kin - min(temp_kin) kin_mat <- kin_mat/max(as.vector(kin_mat))*2 kin_mat_tri <- kin_mat[upper.tri(kin_mat)] kin_mat <- t(kin_mat) kin_mat[upper.tri(kin_mat)] <- kin_mat_tri dir.create(paste0("alphamaterun/unique_", targetdegree, "_", scenario,"_", seed)) write.table(kin_mat, sprintf("./alphamaterun/unique_%i_%i_%i/Nrm.txt", targetdegree,scenario,seed), sep="\t", row.names=TRUE, col.names=FALSE) criterion <- data.frame(Genotype = names(ebv), Value = as.vector(ebv)) write.table(criterion, sprintf("./alphamaterun/unique_%i_%i_%i/Criterion.txt", targetdegree,scenario,seed), sep="\t", row.names=FALSE, quote=FALSE, col.names=FALSE) gender_file = cbind(names(ebv), c(rep(1, pop_size[1]), rep(2, pop_size[2]))) write.table(gender_file, sprintf("./alphamaterun/unique_%i_%i_%i/Gender.txt", targetdegree,scenario,seed), sep="\t", row.names=FALSE, col.names=FALSE, quote = FALSE) spec_text <- sprintf("Seed , %i\nNrmMatrixFile , Nrm.txt\nGenderFile , Gender.txt\nSelCriterionFile , Criterion.txt\nNumberOfMatings , %i\nNumberOfMaleParents , %i\nNumberOfFemaleParents , %i\nTargetDegree , %i\nEqualizeMaleContributions , Yes\nEqualizeFemaleContributions , Yes\nPreselectMales , Yes\nPreselectMalePercentage , 50\nPreselectFemales , Yes\nPreselectFemalePercentage , 50\nEvolAlgNumberOfIterations , 1000\nEvolAlgNumberOfIterationsPrint , 5\nEvolAlgNumberOfIterationsStop , 50\nStop", seed, sum(pop_size), sel_size[1], sel_size[2],targetdegree) write(x = spec_text, file=sprintf("./alphamaterun/unique_%i_%i_%i/AlphaMateSpec.txt", targetdegree, scenario,seed)) sim_path = getwd() setwd(paste0("alphamaterun/unique_", targetdegree, "_", scenario,"_", seed)) suppressWarnings(file.remove("MatingPlanModeOptTarget1.txt")) if(seed<3){ system(sprintf(paste0( sim_path, "/AlphaMate/binaries/AlphaMate_Unix"))) } else{ system(sprintf(paste0( sim_path, "/AlphaMate_0.2.0/bin/AlphaMate_Linux_0.2.02"))) } setwd(sim_path) matings <- read.table(sprintf("./alphamaterun/unique_%i_%i_%i/MatingPlanModeOptTarget1.txt", targetdegree, scenario, seed), sep="", header=TRUE) males = match(as.character(matings$Parent1), names(ebv)[1:pop_size[1]]) females = match(as.character(matings$Parent2), names(ebv)[-(1:pop_size[1])]) sex = rep(0, sum(pop_size)) sex[sample(sum(pop_size), pop_size[2])] = 1 fixed_breeding = cbind(get.ngen(population), 1, males, get.ngen(population),2,females, sex) # Selection based on genomic breeding value population = breeding.diploid(population, fixed.breeding = fixed_breeding, share.genotyped = 1, display.progress = display.progress) }else if( scenario == 239){ targetdegree = 30 temp_geno <- geno pi1 <- rowMeans(temp_geno)/2 temp_kin <- crossprod(temp_geno- 2*pi1) / sum(pi1*(1-pi1)) / 2 selection_index <- as.numeric(ebv) diag_kin <- diag(temp_kin) kin_mat <- temp_kin - min(temp_kin) kin_mat <- kin_mat/max(as.vector(kin_mat))*2 kin_mat_tri <- kin_mat[upper.tri(kin_mat)] kin_mat <- t(kin_mat) kin_mat[upper.tri(kin_mat)] <- kin_mat_tri dir.create(paste0("alphamaterun/unique_", targetdegree, "_", scenario,"_", seed)) write.table(kin_mat, sprintf("./alphamaterun/unique_%i_%i_%i/Nrm.txt", targetdegree,scenario,seed), sep="\t", row.names=TRUE, col.names=FALSE) criterion <- data.frame(Genotype = names(ebv), Value = as.vector(ebv)) write.table(criterion, sprintf("./alphamaterun/unique_%i_%i_%i/Criterion.txt", targetdegree,scenario,seed), sep="\t", row.names=FALSE, quote=FALSE, col.names=FALSE) gender_file = cbind(names(ebv), c(rep(1, pop_size[1]), rep(2, pop_size[2]))) write.table(gender_file, sprintf("./alphamaterun/unique_%i_%i_%i/Gender.txt", targetdegree,scenario,seed), sep="\t", row.names=FALSE, col.names=FALSE, quote = FALSE) spec_text <- sprintf("Seed , %i\nNrmMatrixFile , Nrm.txt\nGenderFile , Gender.txt\nSelCriterionFile , Criterion.txt\nNumberOfMatings , %i\nNumberOfMaleParents , %i\nNumberOfFemaleParents , %i\nTargetDegree , %i\nEqualizeMaleContributions , Yes\nEqualizeFemaleContributions , Yes\nPreselectMales , Yes\nPreselectMalePercentage , 33\nPreselectFemales , Yes\nPreselectFemalePercentage , 33\nEvolAlgNumberOfIterations , 1000\nEvolAlgNumberOfIterationsPrint , 5\nEvolAlgNumberOfIterationsStop , 50\nStop", seed, sum(pop_size), sel_size[1], sel_size[2],targetdegree) write(x = spec_text, file=sprintf("./alphamaterun/unique_%i_%i_%i/AlphaMateSpec.txt", targetdegree, scenario,seed)) sim_path = getwd() setwd(paste0("alphamaterun/unique_", targetdegree, "_", scenario,"_", seed)) suppressWarnings(file.remove("MatingPlanModeOptTarget1.txt")) if(seed<3){ system(sprintf(paste0( sim_path, "/AlphaMate/binaries/AlphaMate_Unix"))) } else{ system(sprintf(paste0( sim_path, "/AlphaMate_0.2.0/bin/AlphaMate_Linux_0.2.02"))) } setwd(sim_path) matings <- read.table(sprintf("./alphamaterun/unique_%i_%i_%i/MatingPlanModeOptTarget1.txt", targetdegree, scenario, seed), sep="", header=TRUE) males = match(as.character(matings$Parent1), names(ebv)[1:pop_size[1]]) females = match(as.character(matings$Parent2), names(ebv)[-(1:pop_size[1])]) sex = rep(0, sum(pop_size)) sex[sample(sum(pop_size), pop_size[2])] = 1 fixed_breeding = cbind(get.ngen(population), 1, males, get.ngen(population),2,females, sex) # Selection based on genomic breeding value population = breeding.diploid(population, fixed.breeding = fixed_breeding, share.genotyped = 1, display.progress = display.progress) }else if( scenario == 240){ targetdegree = 45 temp_geno <- geno pi1 <- rowMeans(temp_geno)/2 temp_kin <- crossprod(temp_geno- 2*pi1) / sum(pi1*(1-pi1)) / 2 selection_index <- as.numeric(ebv) diag_kin <- diag(temp_kin) kin_mat <- temp_kin - min(temp_kin) kin_mat <- kin_mat/max(as.vector(kin_mat))*2 kin_mat_tri <- kin_mat[upper.tri(kin_mat)] kin_mat <- t(kin_mat) kin_mat[upper.tri(kin_mat)] <- kin_mat_tri dir.create(paste0("alphamaterun/unique_", targetdegree, "_", scenario,"_", seed)) write.table(kin_mat, sprintf("./alphamaterun/unique_%i_%i_%i/Nrm.txt", targetdegree,scenario,seed), sep="\t", row.names=TRUE, col.names=FALSE) criterion <- data.frame(Genotype = names(ebv), Value = as.vector(ebv)) write.table(criterion, sprintf("./alphamaterun/unique_%i_%i_%i/Criterion.txt", targetdegree,scenario,seed), sep="\t", row.names=FALSE, quote=FALSE, col.names=FALSE) gender_file = cbind(names(ebv), c(rep(1, pop_size[1]), rep(2, pop_size[2]))) write.table(gender_file, sprintf("./alphamaterun/unique_%i_%i_%i/Gender.txt", targetdegree,scenario,seed), sep="\t", row.names=FALSE, col.names=FALSE, quote = FALSE) spec_text <- sprintf("Seed , %i\nNrmMatrixFile , Nrm.txt\nGenderFile , Gender.txt\nSelCriterionFile , Criterion.txt\nNumberOfMatings , %i\nNumberOfMaleParents , %i\nNumberOfFemaleParents , %i\nTargetDegree , %i\nEqualizeMaleContributions , Yes\nEqualizeFemaleContributions , Yes\nPreselectMales , Yes\nPreselectMalePercentage , 20\nPreselectFemales , Yes\nPreselectFemalePercentage , 20\nEvolAlgNumberOfIterations , 1000\nEvolAlgNumberOfIterationsPrint , 5\nEvolAlgNumberOfIterationsStop , 50\nStop", seed, sum(pop_size), sel_size[1], sel_size[2],targetdegree) write(x = spec_text, file=sprintf("./alphamaterun/unique_%i_%i_%i/AlphaMateSpec.txt", targetdegree, scenario,seed)) sim_path = getwd() setwd(paste0("alphamaterun/unique_", targetdegree, "_", scenario,"_", seed)) suppressWarnings(file.remove("MatingPlanModeOptTarget1.txt")) if(seed<3){ system(sprintf(paste0( sim_path, "/AlphaMate/binaries/AlphaMate_Unix"))) } else{ system(sprintf(paste0( sim_path, "/AlphaMate_0.2.0/bin/AlphaMate_Linux_0.2.02"))) } setwd(sim_path) matings <- read.table(sprintf("./alphamaterun/unique_%i_%i_%i/MatingPlanModeOptTarget1.txt", targetdegree, scenario, seed), sep="", header=TRUE) males = match(as.character(matings$Parent1), names(ebv)[1:pop_size[1]]) females = match(as.character(matings$Parent2), names(ebv)[-(1:pop_size[1])]) sex = rep(0, sum(pop_size)) sex[sample(sum(pop_size), pop_size[2])] = 1 fixed_breeding = cbind(get.ngen(population), 1, males, get.ngen(population),2,females, sex) # Selection based on genomic breeding value population = breeding.diploid(population, fixed.breeding = fixed_breeding, share.genotyped = 1, display.progress = display.progress) } else if( scenario == 247){ # Selection based on genomic breeding value population = breeding.diploid(population, selection.size = c(50,125), breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, display.progress = display.progress) } else if( scenario == 248){ # Selection based on genomic breeding value population = breeding.diploid(population, selection.size = c(45,112), breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, display.progress = display.progress) } else if( scenario == 249){ # Selection based on genomic breeding value population = breeding.diploid(population, selection.size = c(42,105), breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, display.progress = display.progress) } else if( scenario == 250){ # Selection based on genomic breeding value population = breeding.diploid(population, selection.size = c(38,95), breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, display.progress = display.progress) } else if( scenario == 251){ # Selection based on genomic breeding value population = breeding.diploid(population, selection.size = c(36,90), breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, display.progress = display.progress) } else if( scenario == 252){ # Selection based on genomic breeding value population = breeding.diploid(population, selection.size = c(34,85), breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, display.progress = display.progress) } else if( scenario == 253){ # Selection based on genomic breeding value population = breeding.diploid(population, selection.size = c(32,80), breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, display.progress = display.progress) } else if( scenario == 254){ # Selection based on genomic breeding value population = breeding.diploid(population, selection.size = c(30,75), breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, display.progress = display.progress) } else if( scenario == 255){ # Selection based on genomic breeding value population = breeding.diploid(population, selection.size = c(28,70), breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, display.progress = display.progress) } else if(scenario == 303){ weight_rare1 = 1.0 weight_rare2 = 0.0 avoid_mating_inb = 0.5 avoid_mating_kin = 0.95 weight_kinship1 = 0.12 weight_kinship2 = 0.02 ogc_scaler = 1.5 ogc = TRUE sel_size = round(c(40, 100) * 0.32) activ = alpha_hat!= 0 weighting = (1 / freq_alphahat)^(1/3) weighting[weighting==Inf] = 0 weighting[weighting>10] = 10 bve_rare = as.numeric((alpha_hat[activ] * weighting[activ]) %*% geno[activ,]) bve_rare1 = (bve_rare-mean(bve_rare)) /sd(bve_rare) weighting = (1 / freq_alphahat)^(1/2) weighting[weighting==Inf] = 0 weighting[weighting>10] = 10 bve_rare = as.numeric((alpha_hat[activ] * weighting[activ]) %*% geno[activ,]) bve_rare2 = (bve_rare-mean(bve_rare)) /sd(bve_rare) cand = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", export.selected = TRUE, verbose = FALSE) A = kinship.exp(population, gen = get.ngen(population)) diag(A) = (diag(A)-0.5)*2 rel_top = colMeans(A[cand[[1]][,3],]) + colMeans(A[cand[[2]][,3] + pop_size[1],]) unique = -rel_top # bve_kinship1 = (unique - mean(unique))/sd(unique) rel_top = colMeans(A) unique = -rel_top # bve_kinship2 = (unique - mean(unique))/sd(unique) bve_plain = (ebv - mean(ebv))/sd(ebv) new_ebv = weight_rare1 * bve_rare1 + weight_rare2 * bve_rare2 + weight_kinship1 * bve_kinship1 + weight_kinship2 * bve_kinship2 + (1-weight_rare1 - weight_rare2 - weight_kinship1 - weight_kinship2) * bve_plain to_insert = cbind(names(ebv), new_ebv) population = insert.bve(population, bves = to_insert) sel_size_tmp = sel_size sel_size_tmp[1] = pop_size[1] population = breeding.diploid(population, selection.size = sel_size_tmp, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, avoid.mating.inb.quantile = avoid_mating_inb, avoid.mating.kinship.quantile = avoid_mating_kin, avoid.mating.kinship.gen = get.ngen(population), display.progress = display.progress, ogc = ogc, ogc.target = "max.BV", ogc.uniform = "female", ogc.ub.sKin.increase = 0.005 * ogc_scaler, verbose = TRUE) } else if(scenario == 304){ weight_rare1 = 0.4030 weight_rare2 = 0.0020 avoid_mating_inb = 0.900 avoid_mating_kin = 1.000 weight_kinship1 = 0.2030 weight_kinship2 = 0.0290 ogc_scaler = 1.000 ogc = TRUE sel_size = round(c(40, 100) * 0.83) activ = alpha_hat!= 0 weighting = (1 / freq_alphahat)^(1/3) weighting[weighting==Inf] = 0 weighting[weighting>10] = 10 bve_rare = as.numeric((alpha_hat[activ] * weighting[activ]) %*% geno[activ,]) bve_rare1 = (bve_rare-mean(bve_rare)) /sd(bve_rare) weighting = (1 / freq_alphahat)^(1/2) weighting[weighting==Inf] = 0 weighting[weighting>10] = 10 bve_rare = as.numeric((alpha_hat[activ] * weighting[activ]) %*% geno[activ,]) bve_rare2 = (bve_rare-mean(bve_rare)) /sd(bve_rare) cand = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", export.selected = TRUE, verbose = FALSE) A = kinship.exp(population, gen = get.ngen(population)) diag(A) = (diag(A)-0.5)*2 rel_top = colMeans(A[cand[[1]][,3],]) + colMeans(A[cand[[2]][,3] + pop_size[1],]) unique = -rel_top # bve_kinship1 = (unique - mean(unique))/sd(unique) rel_top = colMeans(A) unique = -rel_top # bve_kinship2 = (unique - mean(unique))/sd(unique) bve_plain = (ebv - mean(ebv))/sd(ebv) new_ebv = weight_rare1 * bve_rare1 + weight_rare2 * bve_rare2 + weight_kinship1 * bve_kinship1 + weight_kinship2 * bve_kinship2 + (1-weight_rare1 - weight_rare2 - weight_kinship1 - weight_kinship2) * bve_plain to_insert = cbind(names(ebv), new_ebv) population = insert.bve(population, bves = to_insert) sel_size_tmp = sel_size sel_size_tmp[1] = pop_size[1] population = breeding.diploid(population, selection.size = sel_size_tmp, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, avoid.mating.inb.quantile = avoid_mating_inb, avoid.mating.kinship.quantile = avoid_mating_kin, avoid.mating.kinship.gen = get.ngen(population), display.progress = display.progress, ogc = ogc, ogc.target = "max.BV", ogc.uniform = "female", ogc.ub.sKin.increase = 0.005 * ogc_scaler, verbose = TRUE) } else if(scenario == 320){ activ = alpha_real!= 0 weighting = (1 / freq_alpha)^(1/2) weighting[weighting==Inf] = 0 unique = as.numeric((alpha_real[activ] * weighting[activ]) %*% geno[activ,]) new_ebv = (unique - mean(unique))/sd(unique) population = insert.bve(population, bves = cbind(names(ebv), new_ebv)) population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, display.progress = display.progress) } else if(scenario == 321){ activ = alpha_real!= 0 weighting = (1 / freq_alpha)^(1/2) weighting[weighting==Inf] = 0 weighting[weighting > 5] = 5 unique = as.numeric((alpha_real[activ] * weighting[activ]) %*% geno[activ,]) new_ebv = (unique - mean(unique))/sd(unique) population = insert.bve(population, bves = cbind(names(ebv), new_ebv)) population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, display.progress = display.progress) } else if(scenario == 322){ activ = alpha_real!= 0 weighting = (1 / freq_alpha)^(1/3) weighting[weighting==Inf] = 0 unique = as.numeric((alpha_real[activ] * weighting[activ]) %*% geno[activ,]) new_ebv = (unique - mean(unique))/sd(unique) population = insert.bve(population, bves = cbind(names(ebv), new_ebv)) population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, display.progress = display.progress) } else if(scenario == 323){ activ = alpha_real!= 0 weighting = (1 / freq_alpha)^(1/4) weighting[weighting==Inf] = 0 unique = as.numeric((alpha_real[activ] * weighting[activ]) %*% geno[activ,]) new_ebv = (unique - mean(unique))/sd(unique) population = insert.bve(population, bves = cbind(names(ebv), new_ebv)) population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, display.progress = display.progress) } else if(scenario == 324){ activ = alpha_real!= 0 weighting = 0.8 + 0.2 * (1 / freq_alpha)^(1/2) weighting[weighting==Inf] = 0 unique = as.numeric((alpha_real[activ] * weighting[activ]) %*% geno[activ,]) new_ebv = (unique - mean(unique))/sd(unique) population = insert.bve(population, bves = cbind(names(ebv), new_ebv)) population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, display.progress = display.progress) } else if(scenario == 325){ activ = alpha_real!= 0 weighting = 0.6 + 0.4 * (1 / freq_alpha)^(1/2) weighting[weighting==Inf] = 0 unique = as.numeric((alpha_real[activ] * weighting[activ]) %*% geno[activ,]) new_ebv = (unique - mean(unique))/sd(unique) population = insert.bve(population, bves = cbind(names(ebv), new_ebv)) population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, display.progress = display.progress) } else if(scenario == 326){ activ = alpha_real!= 0 weighting = (pi/2 - asin(sqrt(freq_alpha)))/(sqrt(freq_alpha * (1-freq_alpha))) weighting[weighting==Inf] = 0 unique = as.numeric((alpha_real[activ] * weighting[activ]) %*% geno[activ,]) new_ebv = (unique - mean(unique))/sd(unique) population = insert.bve(population, bves = cbind(names(ebv), new_ebv)) population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, display.progress = display.progress) } else if( scenario == 335){ work_dir = paste0("mixblup_", scenario, "_", seed) if(index == 1){ dir.create(work_dir) system(paste0("cp /home/WUR/pook001/SysDir.inp ", work_dir, "/SysDir.inp")) setwd(work_dir) system(paste0("cp /home/WUR/pook001/msv msv")) system("chmod +x msv") } get.plink(population, gen =bve.gen1, path = "paternal_plink", fam.id = TRUE, type = 1) get.plink(population, gen =bve.gen1, path = "maternal_plink", fam.id = TRUE, type = 2) write.table(file = "EBV.txt", cbind(get.id(population, gen = bve.gen1), ebv), row.names = FALSE, col.names = FALSE, quote = FALSE) write.table(file = "sol.dat", cbind(1,1,1,1:length(alpha_hat), alpha_hat), row.names = FALSE, col.names = FALSE, quote = FALSE) write.table(file = "all_freq", cbind(1:length(alpha_hat), 0), row.names = FALSE, col.names = FALSE, quote = FALSE) system("./msv --matgeno maternal_plink.bed --patgeno paternal_plink.bed --freq all_freq --ebv EBV.txt --sol sol.dat --snpeff 1 --nrgen 2 --selprop 0.1 --bp2M 1e-08") a = read.table("EBV.txt") b = read.table("genetic_MSVs_1gen.txt") c = read.table("Index_5_6.txt") #Piter Index 5/6 d = read.table("ExpBVSelGrOff_per_individual.txt") # Tobias Index write.table(file = "sol.dat", cbind(1,1,1,1:length(alpha_real), alpha_real), row.names = FALSE, col.names = FALSE, quote = FALSE) system("./msv --matgeno maternal_plink.bed --patgeno paternal_plink.bed --freq all_freq --ebv EBV.txt --sol sol.dat --snpeff 1 --nrgen 1 --selprop 0.1 --bp2M 1e-08") a1 = read.table("EBV.txt") b1 = read.table("genetic_MSVs_1gen.txt") c1 = read.table("Index_5_6.txt") #Piter Index 5/6 d1 = read.table("ExpBVSelGrOff_per_individual.txt") # Tobias Index new_ebv = d[,2] insert.bve = cbind(names(ebv), new_ebv) population = insert.bve(population, bves = insert.bve ) acc3 = c(acc3, cor(b[,2],b1[,2])) acc2 = rbind(acc2, t(analyze.bv(population, gen = max(bve.gen))[[1]])) pop1 = breeding.diploid(population, bve = TRUE, bve.gen = bve.gen) acc1 = rbind(acc1, t(analyze.bv(pop1, gen = max(bve.gen))[[1]])) print(colMeans(acc1)) print(colMeans(acc2)) # Selection based on genomic breeding value population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, display.progress = display.progress) } else if( scenario == 336){ work_dir = paste0("mixblup_", scenario, "_", seed) if(index == 1){ dir.create(work_dir) system(paste0("cp /home/WUR/pook001/SysDir.inp ", work_dir, "/SysDir.inp")) setwd(work_dir) system(paste0("cp /home/WUR/pook001/msv msv")) system("chmod +x msv") } get.plink(population, gen =bve.gen1, path = "paternal_plink", fam.id = TRUE, type = 1) get.plink(population, gen =bve.gen1, path = "maternal_plink", fam.id = TRUE, type = 2) write.table(file = "EBV.txt", cbind(get.id(population, gen = bve.gen1), ebv), row.names = FALSE, col.names = FALSE, quote = FALSE) write.table(file = "sol.dat", cbind(1,1,1,1:length(alpha_hat), alpha_hat), row.names = FALSE, col.names = FALSE, quote = FALSE) write.table(file = "all_freq", cbind(1:length(alpha_hat), 0), row.names = FALSE, col.names = FALSE, quote = FALSE) system("./msv --matgeno maternal_plink.bed --patgeno paternal_plink.bed --freq all_freq --ebv EBV.txt --sol sol.dat --snpeff 1 --nrgen 1 --selprop 0.1 --bp2M 1e-08") a = read.table("EBV.txt") b = read.table("genetic_MSVs_1gen.txt") c = read.table("Index_5_6.txt") #Piter Index 5/6 #d = read.table("ExpBVSelGrOff_per_individual.txt") # Tobias Index write.table(file = "sol.dat", cbind(1,1,1,1:length(alpha_real), alpha_real), row.names = FALSE, col.names = FALSE, quote = FALSE) system("./msv --matgeno maternal_plink.bed --patgeno paternal_plink.bed --freq all_freq --ebv EBV.txt --sol sol.dat --snpeff 1 --nrgen 1 --selprop 0.1 --bp2M 1e-08") a1 = read.table("EBV.txt") b1 = read.table("genetic_MSVs_1gen.txt") c1 = read.table("Index_5_6.txt") #Piter Index 5/6 #d1 = read.table("ExpBVSelGrOff_per_individual.txt") # Tobias Index new_ebv = c[,2] insert.bve = cbind(names(ebv), new_ebv) population = insert.bve(population, bves = insert.bve ) acc3 = c(acc3, cor(b[,2],b1[,2])) acc2 = rbind(acc2, t(analyze.bv(population, gen = bve.gen)[[1]])) pop1 = breeding.diploid(population, bve = TRUE, bve.gen = bve.gen) acc1 = rbind(acc1, t(analyze.bv(pop1, gen = bve.gen)[[1]])) print(colMeans(acc1)) print(colMeans(acc2)) # Selection based on genomic breeding value population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, display.progress = display.progress) } else if( scenario == 337){ work_dir = paste0("mixblup_", scenario, "_", seed) if(index == 1){ dir.create(work_dir) system(paste0("cp /home/WUR/pook001/SysDir.inp ", work_dir, "/SysDir.inp")) setwd(work_dir) system(paste0("cp /home/WUR/pook001/msv msv")) system("chmod +x msv") } get.plink(population, gen =bve.gen1, path = "paternal_plink", fam.id = TRUE, type = 1) get.plink(population, gen =bve.gen1, path = "maternal_plink", fam.id = TRUE, type = 2) write.table(file = "EBV.txt", cbind(get.id(population, gen = bve.gen1), ebv), row.names = FALSE, col.names = FALSE, quote = FALSE) write.table(file = "sol.dat", cbind(1,1,1,1:length(alpha_hat), alpha_hat), row.names = FALSE, col.names = FALSE, quote = FALSE) write.table(file = "all_freq", cbind(1:length(alpha_hat), 0), row.names = FALSE, col.names = FALSE, quote = FALSE) system("./msv --matgeno maternal_plink.bed --patgeno paternal_plink.bed --freq all_freq --ebv EBV.txt --sol sol.dat --snpeff 1 --nrgen 1 --selprop 0.1 --bp2M 1e-08") a = read.table("EBV.txt") b = read.table("genetic_MSVs_1gen.txt") c = read.table("Index_5_6.txt") #Piter Index 5/6 #d = read.table("ExpBVSelGrOff_per_individual.txt") # Tobias Index write.table(file = "sol.dat", cbind(1,1,1,1:length(alpha_real), alpha_real), row.names = FALSE, col.names = FALSE, quote = FALSE) system("./msv --matgeno maternal_plink.bed --patgeno paternal_plink.bed --freq all_freq --ebv EBV.txt --sol sol.dat --snpeff 1 --nrgen 1 --selprop 0.1 --bp2M 1e-08") a1 = read.table("EBV.txt") b1 = read.table("genetic_MSVs_1gen.txt") c1 = read.table("Index_5_6.txt") #Piter Index 5/6 #d1 = read.table("ExpBVSelGrOff_per_individual.txt") # Tobias Index new_ebv = c[,3] insert.bve = cbind(names(ebv), new_ebv) population = insert.bve(population, bves = insert.bve ) acc3 = c(acc3, cor(b[,2],b1[,2])) acc2 = rbind(acc2, t(analyze.bv(population, gen = bve.gen)[[1]])) pop1 = breeding.diploid(population, bve = TRUE, bve.gen = bve.gen) acc1 = rbind(acc1, t(analyze.bv(pop1, gen = bve.gen)[[1]])) print(colMeans(acc1)) print(colMeans(acc2)) # Selection based on genomic breeding value population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, display.progress = display.progress) } else if( scenario == 338){ acc2 = rbind(acc2, t(analyze.bv(population, gen = max(bve.gen))[[1]])) # Selection based on genomic breeding value population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, display.progress = display.progress) } else if(scenario == 396){ sel_size_tmp = sel_size # selection on the female side is done without OGC, on male male size selection.size = population size # to provide optiSel with all potential sires # calculation on how to set ub.sKin can be tricky has this corresponds to the increase of inbreeding # that is caused by the male side of the breeding program sel_size_tmp[1] = pop_size[1] # Selection based on genomic breeding value population = breeding.diploid(population, selection.size = sel_size_tmp, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, ogc = TRUE, ogc.target = "max.BV", ogc.uniform = "female", ogc.ub.sKin.increase = 0.0043, display.progress = display.progress) } else if(scenario == 397){ sel_size_tmp = sel_size # selection on the female side is done without OGC, on male male size selection.size = population size # to provide optiSel with all potential sires # calculation on how to set ub.sKin can be tricky has this corresponds to the increase of inbreeding # that is caused by the male side of the breeding program sel_size_tmp[1] = pop_size[1] # Selection based on genomic breeding value population = breeding.diploid(population, selection.size = sel_size_tmp, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, ogc = TRUE, ogc.target = "max.BV", ogc.uniform = "female", ogc.ub.sKin.increase = 0.003, display.progress = display.progress) } else if(scenario == 398){ sel_size_tmp = sel_size # selection on the female side is done without OGC, on male male size selection.size = population size # to provide optiSel with all potential sires # calculation on how to set ub.sKin can be tricky has this corresponds to the increase of inbreeding # that is caused by the male side of the breeding program sel_size_tmp[1] = pop_size[1] # Selection based on genomic breeding value population = breeding.diploid(population, selection.size = sel_size_tmp, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, ogc = TRUE, ogc.target = "max.BV", ogc.uniform = "female", ogc.ub.sKin.increase = 0.004, display.progress = display.progress) } else if(scenario == 399){ sel_size_tmp = sel_size # selection on the female side is done without OGC, on male male size selection.size = population size # to provide optiSel with all potential sires # calculation on how to set ub.sKin can be tricky has this corresponds to the increase of inbreeding # that is caused by the male side of the breeding program sel_size_tmp[1] = pop_size[1] # Selection based on genomic breeding value population = breeding.diploid(population, selection.size = sel_size_tmp, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, ogc = TRUE, ogc.target = "max.BV", ogc.uniform = "female", ogc.ub.sKin.increase = 0.006, display.progress = display.progress) } else if(scenario == 400){ sel_size_tmp = sel_size # selection on the female side is done without OGC, on male male size selection.size = population size # to provide optiSel with all potential sires # calculation on how to set ub.sKin can be tricky has this corresponds to the increase of inbreeding # that is caused by the male side of the breeding program sel_size_tmp[1] = pop_size[1] # Selection based on genomic breeding value population = breeding.diploid(population, selection.size = sel_size_tmp, breeding.size = pop_size, selection.criteria = "bve", share.genotyped = 1, ogc = TRUE, ogc.target = "max.BV", ogc.uniform = "female", ogc.ub.sKin.increase = 0.007, display.progress = display.progress) } } # Generate offspring from the best animals based on traditional EBVs for(index in 1:cycles){ print(index) current_gen = get.ngen(population) - cycles sel_crit = "bve" if(scenario %in% c(320:326)){ sel_crit = "bv" } bve.gen = current_gen if(scenario %in% 335:338){ bve.gen = (current_gen-5):current_gen } population = breeding.diploid(population, bve = TRUE, bve.gen = bve.gen) population = breeding.diploid(population, selection.size = sel_size, breeding.size = pop_size, selection.m.database = cbind(current_gen,1), selection.f.database = cbind(current_gen,2), selection.criteria = sel_crit, share.genotyped = 1, display.progress = display.progress) } #### Evaluation of the population #### Evaluation of the population ngen = get.ngen(population) bv = numeric(ngen) bv_sd = numeric(ngen) inb = numeric(ngen) share_fixed = numeric(ngen) share_fixed_qtl = numeric(ngen) share_fixed_good = numeric(ngen) share_fixed_bad = numeric(ngen) for(index in 1:ngen){ print(index) bv[index] = mean(get.bv(population, gen = index)) bv_sd[index] = sd(get.bv(population, gen = index)) inb[index] = mean(inbreeding.emp(population, gen = index)) geno = get.geno(population, gen = index) p = rowMeans(geno)/2 share_fixed[index] = mean( p ==0 | p ==1) share_fixed_qtl[index] = mean( (p ==0 | p ==1)[population$info$effect.p]) share_fixed_good[index] = mean(((alpha_real > 0 & p ==1) | (alpha_real < 0 & p ==0))[population$info$effect.p]) share_fixed_bad[index] = mean(((alpha_real < 0 & p ==1) | (alpha_real > 0 & p ==0))[population$info$effect.p]) } save(file = title, list = c("bv", "bv_sd", "inb", "share_fixed", "share_fixed_qtl", "share_fixed_good", "share_fixed_bad"))