SIMER

Data Simulation for Life Science and Breeding

Authors:

Design and Maintenance: Dong Yin, Xuanning Zhang, Lilin Yin ,Haohao Zhang, and Xiaolei Liu.
Contributors: Zhenshuang Tang, Jingya Xu, Xiaohui Yuan, Xiang Zhou, Xinyun Li, and Shuhong Zhao.

If you have any bug reports or questions, please feed back :point_right:here:point_left:.

:toolbox: Relevant software tools for genetic analyses and genomic breeding

Contents

• Installation
• Data Preparation
  • Genotype
  • Genetic map
  • Pedigree
• Data Input
  • Basic
  • Optional
• Quick Start
  • Quick Start for Population Simulation
  • Quick Start for Genotype Simulation
  • Quick Start for Phenotype Simulation
• Genotype Simulation
  • Gallery of genotype simulation parameters
  • Generate an external or species-specific or random genetic map
  • Generate an external or species-specific or random genotype matrix
  • Generate a genotype matrix with complete linkage disequilibrium
  • Add chromosome crossovers and mutations to genotype matrix
• Phenotype Simulation
  • Gallery of phenotype simulation parameters
  • Generate phenotype using an external or species-specific or random genotype matrix
  • Generate continuous phenotype
  • Generate case-control phenotype
  • Generate categorical phenotype
  • Generate phenotype using A model
  • Generate phenotype using AD model
  • Generate phenotype using GxG model
  • Generate phenotype using Repeated Record model
  • Generate phenotype controlled by QTNs subject to Normal distribution
  • Generate phenotype controlled by QTNs subject to Geometric distribution
  • Generate phenotype controlled by QTNs subject to Gamma distribution
  • Generate phenotype controlled by QTNs subject to Beta distribution
  • Generate phenotype with fixed effect and covariate and environmental random effect
  • Generate phenotype using GxE model
  • Generate phenotype using ExE model
  • Generate phenotype controlled by varied QTN effect distribution
• Population Simulation of Multiple-Generation with Genotype and Phenotype
  • Gallery of population simulation parameters
  • Individual selection for a single trait
  • Family selection for a single trait
  • Within-family selection for a single trait
  • Combined selection for a single trait
  • Tandem selection for multiple traits
  • Independent culling selection for multiple traits
  • Index selection for multiple traits
  • Clone for plants
  • Doubled haploid for plants
  • Self-pollination for plants and micro-organisms
  • Random mating for plants and animals
  • Random mating excluding self-pollination for animals
  • Assortative mating for plants and animals
  • Disassortative mating for plants and animals
  • Two way cross for animals
  • Three way cross for animals
  • Four way cross for animals
  • Back cross for animals
  • User-designed pedigree mating for plants and animals
  • AN EASY WAY TO GENERATE A POPULATION
• Breeding Program Design
  • Gallery of breeding program design parameters
  • Preparation of a breeding program design
  • Evaluation of a breeding program design
• Global Options
  • Gallery of global parameters
  • Counts of total population size
  • Multi-thread computation
  • Multi-population simulation
  • File output
  • Generation-selective output
• Output
  • Annotation data
  • Genotype data
  • Phenotype data
• Citation
• FAQ and Hints

Installation

back to top

WE STRONGLY RECOMMEND TO INSTALL SIMER ON Microsoft R Open (https://mran.microsoft.com/download/).

Installation

• The stable version:

install.packages("simer")

• The latest version:

devtools::install_github("xiaolei-lab/SIMER")

After installed successfully, SIMER can be loaded by typing

> library(simer)

Typing ?simer could get the details of all parameters.

Data Preparation

Genotype

back to top

Genotype data should be Numeric format (m rows and n columns, m is the number of SNPs, n is the number of individuals). Other genotype data, such as PLINK Binary format (details see http://zzz.bwh.harvard.edu/plink/data.shtml#bed), VCF, or Hapmap can be converted to Numeric format using MVP.Data function in the rMVP (https://github.com/xiaolei-lab/rMVP).

genotype.txt

Genetic map

back to top
A genetic map is necessary in SIMER. The first column is the SNP name, the second column is the Chromosome ID, the third column is physical position, the fourth column is REF, and the fifth column is ALT. This will be used to generate annotation data, genotype data, and phenotype data.

map.txt

Pedigree

back to top
SIMER supports user designed pedigree to control mating process. User designed pedigree is useful only in userped reproduction. The first column is sample id, the second column is paternal id, and the third column is maternal id. Please make sure that paternal id and maternal id can match to genotype data.

userped.txt

Data Input

Basic

back to top
At least users should prepare two datasets: genotypic map and genotype data.

genotype data, Numeric format (m rows and n columns, m is the number of SNPs, n is the number of individuals)
genotypic map, SNP map information, the first column is SNP name, the second column is Chromosome ID, the third column is physical position, the fourth column is REF, and the fifth column is ALT.

pop.geno <- read.table("genotype.txt")
pop.map <- read.table("map.txt" , head = TRUE)

Optional

back to top
The mating process can be designed by user-designed pedigree.

pedigree, pedigree information, the first column is sample id, the second column is paternal id, and the third column is maternal id. Note that the individuals in the pedigree do not need to be sorted by the date of birth, and the missing value can be replaced by NA or 0.

userped <- read.table("userped.txt", header = TRUE)

Quick Start

back to top

All simulation processes can be divided into two steps: 1) generation of simulation parameters; 2) run simulation process.

Quick Start for Population Simulation

back to top

A quick start for Population Simulation is shown below:

# Generate all simulation parameters
SP <- param.simer(out = "simer")

# Run Simer
SP <- simer(SP)

Quick Start for Genotype Simulation

back to top

A quick start for Genotype Simulation is shown below:

# Generate annotation simulation parameters
SP <- param.annot(species = "pig")
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2)

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)

Quick Start for Phenotype Simulation

back to top

A quick start for Phenotype Simulation is shown below:

# Generate annotation simulation parameters
SP <- param.annot(species = "pig")
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2)
# Generate phenotype simulation parameters
SP <- param.pheno(SP = SP, phe.h2A = list(tr1 = 0.3))

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)

Genotype Simulation

back to top

Genotype data in SIMER is generated randomly or through an external genotype matrix. Chromosome crossovers and base mutations depend on block information and recombination information of Annotation data.

Gallery of genotype simulation parameters

back to top

genotype, main function of Genotype Simulation:

annotation, main function of Annotation Simulation:

Generate an external or species-specific or random genetic map

back to top

Users can generate a genetic map by inputting an external genetic map.

# Real genotypic map
mapPath <- system.file("extdata", "06map", "pig_map.txt", package = "simer")
pop.map <- read.table(mapPath, header = TRUE)

# Generate annotation simulation parameters
SP <- param.annot(pop.map = pop.map)

# Run annotation simulation
SP <- annotation(SP)

Users can also use the inner real genetic map with species, which can be "arabidopsis", "cattle", "chicken", "dog", "horse", "human", "maize", "mice", "pig", and "rice".

# Generate annotation simulation parameters
SP <- param.annot(species = "pig")

# Run annotation simulation
SP <- annotation(SP)

Users can generate a random genetic map with pop.marker, num.chr, and len.chr.

# Generate annotation simulation parameters
SP <- param.annot(pop.marker = 1e4, num.chr = 18, len.chr = 1.5e8)

# Run annotation simulation
SP <- annotation(SP)

Generate an external or species-specific or random genotype matrix

back to top

Users can use real genotype data with specific genetic structure for subsequent simulation.

# Create a genotype matrix
# pop.geno <- read.table("genotype.txt")
# pop.geno <- bigmemory::attach.big.matrix("genotype.geno.desc")
pop.geno <- matrix(c(0, 1, 2, 0), nrow = 1e4, ncol = 1e2, byrow = TRUE)

# Generate annotation simulation parameters
SP <- param.annot(pop.marker = 1e4)
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.geno = pop.geno)

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)

Users can also generate genotype matrix with the inner real genetic map with species, which can be "arabidopsis", "cattle", "chicken", "dog", "horse", "human", "maize", "mice", "pig", and "rice".

# Generate annotation simulation parameters
SP <- param.annot(species = "pig")
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2)

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)

Users can also specify pop.marker and pop.ind to generate random genotype data.

# Generate annotation simulation parameters
SP <- param.annot(pop.marker = 1e4)
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2)

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)

Generate a genotype matrix with complete linkage disequilibrium

back to top

Users can generate a genotype matrix with complete linkage disequilibrium by incols = 2 and cld = TRUE.

# Generate annotation simulation parameters
SP <- param.annot(pop.marker = 1e4)
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2, incols = 2, cld = TRUE)

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)

Add chromosome crossovers and mutations to genotype matrix

back to top

With annotation data, chromosome crossovers and mutations can be added to a genotype matrix.

# Generate annotation simulation parameters
# If recom.spot = TRUE, chromsome crossovers will be added to genotype matrix
SP <- param.annot(pop.marker = 1e4, recom.spot = TRUE)
# Generate genotype simulation parameters
# Base mutation rate of QTN and SNP are 1e8
SP <- param.geno(SP = SP, pop.ind = 1e2, rate.mut = list(qtn = 1e-8, snp = 1e-8))

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)

Note that recombination only exists in meiosis. Therefore, some reproduction methods such as clone do not have recombination processes. Users can set recom.spot = FALSE to add only mutations to the genotype matrix.

# Generate annotation simulation parameters
# If recom.spot = FALSE, chromsome crossovers will not be added to genotype matrix
SP <- param.annot(pop.marker = 1e4, recom.spot = FALSE)
# Generate genotype simulation parameters
# Base mutation rate of QTN and SNP are 1e8
SP <- param.geno(SP = SP, pop.ind = 1e2, rate.mut = list(qtn = 1e-8, snp = 1e-8))

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)

Phenotype Simulation

back to top

Phenotype data in SIMER is generated according to different models, which include:
(1) Single-Trait Model
(2) Multiple-Trait Model
(3) Repeated Record Model
(4) Genetic Effect Model (Additive effect, Dominant effect, and Genetic-Genetic interaction effect)
(5) Genetic Model with Varied QTN Effect Distributions (QTN effect distribution: Normal distribution, Geometric distribution, Gamma distribution, Beta distribution, and their combination)
(6) Linear Mixed Model (Fixed effect, Covariate, Environmental Random effect, Genetic Random effect, Genetic-Environmental interaction effect, and Environmental-Environmental interaction effect)

Gallery of phenotype simulation parameters

back to top

phenotype, main function of Phenotype Simulation:

annotation, main function of Annotation Simulation:

Generate phenotype using an external or species-specific or random genotype matrix

back to top

Users can use real genotype data with specific genetic structure to generate phenotype.

# Create a genotype matrix
# pop.geno <- read.table("genotype.txt")
# pop.geno <- bigmemory::attach.big.matrix("genotype.geno.desc")
pop.geno <- matrix(c(0, 1, 2, 0), nrow = 1e4, ncol = 1e2, byrow = TRUE)

# Generate annotation simulation parameters
SP <- param.annot(pop.marker = 1e4)
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.geno = pop.geno)
# Generate phenotype simulation parameters
SP <- param.pheno(SP = SP, phe.h2A = list(tr1 = 0.3))

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)

Users can also generate phenotype using species-specific genotype matrix.

# Generate annotation simulation parameters
SP <- param.annot(species = "pig")
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2)
# Generate phenotype simulation parameters
SP <- param.pheno(SP = SP, phe.h2A = list(tr1 = 0.3))

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)

Users can also generate phenotype using random genotype.

# Generate annotation simulation parameters
SP <- param.annot(pop.marker = 1e4)
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2)
# Generate phenotype simulation parameters
SP <- param.pheno(SP = SP, phe.h2A = list(tr1 = 0.3))

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)

Generate continuous phenotype

back to top

SIMER generates continuous phenotypes by default. Continuous phenotype simulation is displayed as follows:
If users want to output files, please see File output.

# Generate annotation simulation parameters
SP <- param.annot(pop.marker = 1e4, qtn.num = list(tr1 = 10), qtn.model = "A") # Additive effect
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2) # random genotype
# Generate phenotype simulation parameters
SP <- param.pheno(
  SP = SP,
  phe.type = list(tr1 = "continuous"),
  phe.model = list(tr1 = "T1 = A + E"), # "T1" (Trait 1) consists of Additive effect and Residual effect
  # phe.var = list(tr1 = 100),
  phe.h2A = list(tr1 = 0.3)
)

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)

Multiple-trait simulation of continuous phenotype is displayed as follows:
If users want to output files, please see File output.

# Generate annotation simulation parameters
SP <- param.annot(pop.marker = 1e4, qtn.num = list(tr1 = 10, tr2 = 10), qtn.model = "A") # Additive effect
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2) # random genotype
# Generate phenotype simulation parameters
SP <- param.pheno(
  SP = SP,
  phe.type = list(tr1 = "continuous", tr2 = "continuous"),
  phe.model = list(
    tr1 = "T1 = A + E", # "T1" (Trait 1) consists of Additive effect and Residual effect
    tr2 = "T2 = A + E"  # "T2" (Trait 2) consists of Additive effect and Residual effect
  ),
  # phe.var = list(tr1 = 100, tr2 = 100),
  phe.h2A = list(tr1 = 0.3, tr2 = 0.3),
  phe.corA = matrix(c(1, 0.5, 0.5, 1), 2, 2) # Additive genetic correlation
)

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)

Generate case-control phenotype

back to top

SIMER generates case-control phenotypes by phe.type. phe.type consists of the variable names and their percentages. Case-control phenotype simulation is displayed as follows:
If users want to output files, please see File output.

# Generate annotation simulation parameters
SP <- param.annot(pop.marker = 1e4, qtn.num = list(tr1 = 10), qtn.model = "A") # Additive effect
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2) # random genotype
# Generate phenotype simulation parameters
SP <- param.pheno(
  SP = SP,
  phe.type = list(tr1 = list(case = 0.01, control = 0.99)), # "T1" (Trait 1) consists of 1% case and 99% control
  phe.model = list(tr1 = "T1 = A + E"), # "T1" (Trait 1) consists of Additive effect and Residual effect
  # phe.var = list(tr1 = 100),
  phe.h2A = list(tr1 = 0.3)
)

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)

Multiple-trait simulation of case-control phenotype is displayed as follows:
If users want to output files, please see File output.

# Generate annotation simulation parameters
SP <- param.annot(pop.marker = 1e4, qtn.num = list(tr1 = 10, tr2 = 10), qtn.model = "A") # Additive effect
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2) # random genotype
# Generate phenotype simulation parameters
SP <- param.pheno(
  SP = SP,
  phe.type = list(
    tr1 = list(case = 0.01, control = 0.99), # "T1" (Trait 1) consists of 1% case and 99% control
    tr2 = list(case = 0.01, control = 0.99)  # "T2" (Trait 2) consists of 1% case and 99% control
   ),
  phe.model = list(
    tr1 = "T1 = A + E", # "T1" (Trait 1) consists of Additive effect and Residual effect
    tr2 = "T2 = A + E"  # "T2" (Trait 2) consists of Additive effect and Residual effect
  ),
  # phe.var = list(tr1 = 100, tr2 = 100),
  phe.h2A = list(tr1 = 0.3, tr2 = 0.3),
  phe.corA = matrix(c(1, 0.5, 0.5, 1), 2, 2) # Additive genetic correlation
)

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)

Generate categorical phenotype

back to top

SIMER generates categorical phenotypes by phe.type. phe.type consists of the variable names and their percentages. Categorical phenotype simulation is displayed as follows:
If users want to output files, please see File output.

# Generate annotation simulation parameters
SP <- param.annot(pop.marker = 1e4, qtn.num = list(tr1 = 10), qtn.model = "A") # Additive effect
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2) # random genotype
# Generate phenotype simulation parameters
SP <- param.pheno(
  SP = SP,
  phe.type = list(tr1 = list(low = 0.3, medium = 0.4, high = 0.3)), # "T1" (Trait 1) consists of 30% low, 40% medium, and 30% high
  phe.model = list(tr1 = "T1 = A + E"), # "T1" (Trait 1) consists of Additive effect and Residual effect
  # phe.var = list(tr1 = 100),
  phe.h2A = list(tr1 = 0.3)
)

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)

Multiple-trait simulation of categorical phenotype is displayed as follows:
If users want to output files, please see File output.

# Generate annotation simulation parameters
SP <- param.annot(pop.marker = 1e4, qtn.num = list(tr1 = 10, tr2 = 10), qtn.model = "A") # Additive effect
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2) # random genotype
# Generate phenotype simulation parameters
SP <- param.pheno(
  SP = SP,
  phe.type = list(
    tr1 = list(low = 0.3, medium = 0.4, high = 0.3), # "T1" (Trait 1) consists of 30% low, 40% medium, and 30% high
    tr2 = list(low = 0.3, medium = 0.4, high = 0.3)  # "T2" (Trait 2) consists of 30% low, 40% medium, and 30% high
   ),
  phe.model = list(
    tr1 = "T1 = A + E", # "T1" (Trait 1) consists of Additive effect and Residual effect
    tr2 = "T2 = A + E"  # "T2" (Trait 2) consists of Additive effect and Residual effect
  ),
  # phe.var = list(tr1 = 100, tr2 = 100),
  phe.h2A = list(tr1 = 0.3, tr2 = 0.3),
  phe.corA = matrix(c(1, 0.5, 0.5, 1), 2, 2) # Additive genetic correlation
)

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)

Generate phenotype using A model

back to top

In an "A" model, SIMER only considers an Additive effect as a genetic effect. Users should prepare Additive QTN effect in the Annotation data to generate an Additive Individual effect. An Additive single-trait simulation is displayed as follows:
If users want to output files, please see File output.

# Generate annotation simulation parameters
SP <- param.annot(pop.marker = 1e4, qtn.num = list(tr1 = 10), qtn.model = "A") # Additive effect
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2) # random genotype
# Generate phenotype simulation parameters
SP <- param.pheno(
  SP = SP,
  phe.model = list(tr1 = "T1 = A + E"), # "T1" (Trait 1) consists of Additive effect and Residual effect
  # phe.var = list(tr1 = 100),
  phe.h2A = list(tr1 = 0.3)
)

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)

In the multiple-trait simulation, SIMER builds accurate Additive genetic correlation among multiple traits. An Additive multiple-trait simulation is displayed as follows:
If users want to output files, please see File output.

# Generate annotation simulation parameters
SP <- param.annot(pop.marker = 1e4, qtn.num = list(tr1 = 10, tr2 = 10), qtn.model = "A") # Additive effect
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2) # random genotype
# Generate phenotype simulation parameters
SP <- param.pheno(
  SP = SP,
  phe.model = list(
    tr1 = "T1 = A + E", # "T1" (Trait 1) consists of Additive effect and Residual effect
    tr2 = "T2 = A + E"  # "T2" (Trait 2) consists of Additive effect and Residual effect
  ),
  # phe.var = list(tr1 = 100, tr2 = 100),
  phe.h2A = list(tr1 = 0.3, tr2 = 0.3),
  phe.corA = matrix(c(1, 0.5, 0.5, 1), 2, 2) # Additive genetic correlation
)

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)

Generate phenotype using AD model

back to top

In an "AD" model, SIMER considers Additive effect and Dominant effect as genetic effect. Users should prepare Additive QTN effect and Dominant QTN effect in the Annotation data to generate an Additive Individual effect and Dominant Individual effect. Additive and Dominant single-trait simulation is displayed as follows:
If users want to output files, please see File output.

# Generate annotation simulation parameters
SP <- param.annot(pop.marker = 1e4, qtn.num = list(tr1 = 10), qtn.model = "A + D") # Additive effect and Dominant effect
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2) # random genotype
# Generate phenotype simulation parameters
SP <- param.pheno(
  SP = SP,
  phe.model = list(tr1 = "T1 = A + D + E"), # "T1" (Trait 1) consists of Additive effect, Dominant effect, and Residual effect
  # phe.var = list(tr1 = 100),
  phe.h2A = list(tr1 = 0.3),
  phe.h2D = list(tr1 = 0.1)
)

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)

In multiple-trait simulation, SIMER builds accurate Additive genetic correlation and accurate Dominant genetic correlation among multiple traits. An Additive and Dominant multiple-trait simulation is displayed as follows:
If users want to output files, please see File output.

# Generate annotation simulation parameters
SP <- param.annot(pop.marker = 1e4, qtn.num = list(tr1 = 10, tr2 = 10), qtn.model = "A + D") # Additive effect and Dominant effect
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2) # random genotype
# Generate phenotype simulation parameters
SP <- param.pheno(
  SP = SP,
  phe.model = list(
    tr1 = "T1 = A + D + E", # "T1" (Trait 1) consists of Additive effect, Dominant effect, and Residual effect
    tr2 = "T2 = A + D + E"  # "T2" (Trait 2) consists of Additive effect, Dominant effect, and Residual effect
  ),
  # phe.var = list(tr1 = 100, tr2 = 100),
  phe.h2A = list(tr1 = 0.3, tr2 = 0.3),
  phe.h2D = list(tr1 = 0.1, tr2 = 0.1),
  phe.corA = matrix(c(1, 0.5, 0.5, 1), 2, 2), # Additive genetic correlation
  phe.corD = matrix(c(1, 0.5, 0.5, 1), 2, 2)  # Dominant genetic correlation
)

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)

Generate phenotype using GxG model

back to top

In a "GxG" model, SIMER considers Genetic-Genetic effect as a genetic effect. Users should prepare Genetic-Genetic QTN effect in the Annotation data to generate Genetic-Genetic Individual effect. An example of Additive-Dominant interaction in single-trait simulation is displayed as follows:
If users want to output files, please see File output.

# Generate annotation simulation parameters
SP <- param.annot(pop.marker = 1e4, qtn.num = list(tr1 = 10), qtn.model = "A + D + A:D") # Additive effect, Dominant effect, and Additive-Dominant interaction effect
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2) # random genotype
# Generate phenotype simulation parameters
SP <- param.pheno(
  SP = SP,
  phe.model = list(tr1 = "T1 = A + D + A:D + E"), # "T1" (Trait 1) consists of Additive effect, Dominant effect, Additive-Dominant interaction effect, and Residual effect
  # phe.var = list(tr1 = 100),
  phe.h2A = list(tr1 = 0.3),
  phe.h2D = list(tr1 = 0.1),
  phe.h2GxG = list(tr1 = list("A:D" = 0.1))
)

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)

In the multiple-trait simulation, SIMER builds accurate Genetic-Genetic interaction correlation among multiple traits. An example of Additive-Dominant interaction in multiple-trait simulation is displayed as follows:
If users want to output files, please see File output.

# Generate annotation simulation parameters
SP <- param.annot(pop.marker = 1e4, qtn.num = list(tr1 = 10, tr2 = 10), qtn.model = "A + D + A:D") # Additive effect, Dominant effect, and Additive-Dominant interaction effect
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2) # random genotype
# Generate phenotype simulation parameters
SP <- param.pheno(
  SP = SP,
  phe.model = list(
    tr1 = "T1 = A + D + A:D + E", # "T1" (Trait 1) consists of Additive effect, Dominant effect, Additive-Dominant interaction effect, and Residual effect
    tr2 = "T2 = A + D + A:D + E"  # "T2" (Trait 2) consists of Additive effect, Dominant effect, Additive-Dominant interaction effect, and Residual effect
  ),
  # phe.var = list(tr1 = 100, tr2 = 100),
  phe.h2A = list(tr1 = 0.3, tr2 = 0.3),
  phe.h2D = list(tr1 = 0.1, tr2 = 0.1),
  phe.h2GxG = list(tr1 = list("A:D" = 0.1), tr2 = list("A:D" = 0.1)),
  phe.corA = matrix(c(1, 0.5, 0.5, 1), 2, 2),                 # Additive genetic correlation
  phe.corD = matrix(c(1, 0.5, 0.5, 1), 2, 2),                 # Dominant genetic correlation
  phe.corGxG = list("A:D" = matrix(c(1, 0.5, 0.5, 1), 2, 2))  # Additive-Dominant interaction genetic correlation
)

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)

Generate phenotype using Repeated Record model

back to top

In the Repeated Record model, SIMER adds a PE (Permanent Environmental) effect to the phenotype. The number of repeated records can be set by pop.rep. In the meantime, pop.rep.bal can be used to determine whether repeated records are balanced. The Repeated Record in a single-trait simulation is displayed as follows:
If users want to output files, please see File output.

# Generate annotation simulation parameters
SP <- param.annot(pop.marker = 1e4, qtn.num = list(tr1 = 10), qtn.model = "A") # Additive effect
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2) # random genotype
# Generate phenotype simulation parameters
SP <- param.pheno(
  SP = SP,
  pop.rep = 2,                          # The number of repeated records is 2
  pop.rep.bal = TRUE,                   # Repeated records are balanced
  phe.model = list(tr1 = "T1 = A + E"), # "T1" (Trait 1) consists of Additive effect and Residual effect
  # phe.var = list(tr1 = 100),
  phe.h2A = list(tr1 = 0.3)
)

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)

In the multiple-trait simulation, SIMER builds accurate Permanent Environmental correlation among multiple traits. Repeated Record in multiple-trait simulation is displayed as follows:
If users want to output files, please see File output.

# Generate annotation simulation parameters
SP <- param.annot(pop.marker = 1e4, qtn.num = list(tr1 = 10, tr2 = 10), qtn.model = "A") # Additive effect
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2) # random genotype
# Generate phenotype simulation parameters
SP <- param.pheno(
  SP = SP,
  pop.rep = 2,          # The number of repeated records is 2
  pop.rep.bal = TRUE,   # Repeated records are balanced
  phe.model = list(
    tr1 = "T1 = A + E", # "T1" (Trait 1) consists of Additive effect and Residual effect
    tr2 = "T2 = A + E"  # "T2" (Trait 2) consists of Additive effect and Residual effect
  ),
  # phe.var = list(tr1 = 100, tr2 = 100),
  phe.h2A = list(tr1 = 0.3, tr2 = 0.3),
  phe.corA = matrix(c(1, 0.5, 0.5, 1), 2, 2), # Additive genetic correlation
  phe.corPE = matrix(c(1, 0.5, 0.5, 1), 2, 2) # Permanent Environmental correlation
)

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)

Generate phenotype controlled by QTNs subject to Normal distribution

back to top

Normal distribution is the most common QTN effect distribution. Phenotype controlled by QTNs subject to Normal distribution in single-trait simulation is displayed as follows:
If users want to output files, please see File output.

# Generate annotation simulation parameters
SP <- param.annot(
  pop.marker = 1e4,
  qtn.num = list(tr1 = 10),
  qtn.model = "A",
  qtn.dist = list(tr1 = "norm"),
  qtn.var = list(tr1 = 1)
)
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2) # random genotype
# Generate phenotype simulation parameters
SP <- param.pheno(
  SP = SP,
  phe.model = list(tr1 = "T1 = A + E"), # "T1" (Trait 1) consists of Additive effect and Residual effect
  # phe.var = list(tr1 = 100),
  phe.h2A = list(tr1 = 0.3)
)

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)

Phenotype controlled by QTNs subject to Normal distribution in multiple-trait simulation is displayed as follows:
If users want to output files, please see File output.

# Generate annotation simulation parameters
SP <- param.annot(
  pop.marker = 1e4,
  qtn.num = list(tr1 = 10, tr2 = 10),
  qtn.model = "A",
  qtn.dist = list(tr1 = "norm", tr2 = "norm"),
  qtn.var = list(tr1 = 1, tr2 = 1)
)
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2) # random genotype
# Generate phenotype simulation parameters
SP <- param.pheno(
  SP = SP,
  phe.model = list(
    tr1 = "T1 = A + E", # "T1" (Trait 1) consists of Additive effect and Residual effect
    tr2 = "T2 = A + E"  # "T2" (Trait 2) consists of Additive effect and Residual effect
  ),
  # phe.var = list(tr1 = 100, tr2 = 100),
  phe.h2A = list(tr1 = 0.3, tr2 = 0.3),
  phe.corA = matrix(c(1, 0.5, 0.5, 1), 2, 2) # Additive genetic correlation
)

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)

Generate phenotype controlled by QTNs subject to Geometric distribution

back to top

Geometric distribution is the probability of success for the first time obtained only after K trials among the N Bernoulli trials. Geometric distribution can be used as a QTN effect distribution. Phenotype controlled by QTNs subject to Geometric distribution in single-trait simulation is displayed as follows:
If users want to output files, please see File output.

# Generate annotation simulation parameters
SP <- param.annot(
  pop.marker = 1e4,
  qtn.num = list(tr1 = 10),
  qtn.model = "A",
  qtn.dist = list(tr1 = "geom"),
  qtn.prob = list(tr1 = 0.5)
)
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2) # random genotype
# Generate phenotype simulation parameters
SP <- param.pheno(
  SP = SP,
  phe.model = list(tr1 = "T1 = A + E"), # "T1" (Trait 1) consists of Additive effect and Residual effect
  # phe.var = list(tr1 = 100),
  phe.h2A = list(tr1 = 0.3)
)

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)

Phenotype controlled by QTNs subject to Geometric distribution in multiple-trait simulation is displayed as follows:
If users want to output files, please see File output.

# Generate annotation simulation parameters
SP <- param.annot(
  pop.marker = 1e4,
  qtn.num = list(tr1 = 10, tr2 = 10),
  qtn.model = "A",
  qtn.dist = list(tr1 = "geom", tr2 = "geom"),
  qtn.prob = list(tr1 = 0.5, tr2 = 0.5)
)
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2) # random genotype
# Generate phenotype simulation parameters
SP <- param.pheno(
  SP = SP,
  phe.model = list(
    tr1 = "T1 = A + E", # "T1" (Trait 1) consists of Additive effect and Residual effect
    tr2 = "T2 = A + E"  # "T2" (Trait 2) consists of Additive effect and Residual effect
  ),
  # phe.var = list(tr1 = 100, tr2 = 100),
  phe.h2A = list(tr1 = 0.3, tr2 = 0.3),
  phe.corA = matrix(c(1, 0.5, 0.5, 1), 2, 2) # Additive genetic correlation
)

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)

Generate phenotype controlled by QTNs subject to Gamma distribution

back to top

Gamma distribution is the sum of N independent exponential random variables. Note that Exponential distribution is a special form of Gamma distribution when qtn.shape = 1 and qtn.scale = 1. Phenotype controlled by QTNs subject to Gamma distribution in single-trait simulation is displayed as follows:
If users want to output files, please see File output.

# Generate annotation simulation parameters
SP <- param.annot(
  pop.marker = 1e4,
  qtn.num = list(tr1 = 10),
  qtn.model = "A",
  qtn.dist = list(tr1 = "gamma"),
  qtn.shape = list(tr1 = 1),
  qtn.scale = list(tr1 = 1)
)
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2) # random genotype
# Generate phenotype simulation parameters
SP <- param.pheno(
  SP = SP,
  phe.model = list(tr1 = "T1 = A + E"), # "T1" (Trait 1) consists of Additive effect and Residual effect
  # phe.var = list(tr1 = 100),
  phe.h2A = list(tr1 = 0.3)
)

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)

Phenotype controlled by QTNs subject to Gamma distribution in multiple-trait simulation is displayed as follows:
If users want to output files, please see File output.

# Generate annotation simulation parameters
SP <- param.annot(
  pop.marker = 1e4,
  qtn.num = list(tr1 = 10, tr2 = 10),
  qtn.model = "A",
  qtn.dist = list(tr1 = "gamma", tr2 = "gamma"),
  qtn.shape = list(tr1 = 1, tr2 = 1),
  qtn.scale = list(tr1 = 1, tr2 = 1)
)
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2) # random genotype
# Generate phenotype simulation parameters
SP <- param.pheno(
  SP = SP,
  phe.model = list(
    tr1 = "T1 = A + E", # "T1" (Trait 1) consists of Additive effect and Residual effect
    tr2 = "T2 = A + E"  # "T2" (Trait 2) consists of Additive effect and Residual effect
  ),
  # phe.var = list(tr1 = 100, tr2 = 100),
  phe.h2A = list(tr1 = 0.3, tr2 = 0.3),
  phe.corA = matrix(c(1, 0.5, 0.5, 1), 2, 2) # Additive genetic correlation
)

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)

Generate phenotype controlled by QTNs subject to Beta distribution

back to top

Beta distribution is a density function of conjugate prior distribution as Bernoulli distribution and Binomial distribution. Phenotype controlled by QTNs subject to the Beta distribution in single-trait simulation is displayed as follows:
If users want to output files, please see File output.

# Generate annotation simulation parameters
SP <- param.annot(
  pop.marker = 1e4,
  qtn.num = list(tr1 = 10),
  qtn.model = "A",
  qtn.dist = list(tr1 = "beta"),
  qtn.shape1 = list(tr1 = 1),
  qtn.shape2 = list(tr1 = 1),
  qtn.ncp = list(tr1 = 0)
)
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2) # random genotype
# Generate phenotype simulation parameters
SP <- param.pheno(
  SP = SP,
  phe.model = list(tr1 = "T1 = A + E"), # "T1" (Trait 1) consists of Additive effect and Residual effect
  # phe.var = list(tr1 = 100),
  phe.h2A = list(tr1 = 0.3)
)

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)

Phenotype controlled by QTNs subject to Beta distribution in multiple-trait simulation is displayed as follows:
If users want to output files, please see File output.

# Generate annotation simulation parameters
SP <- param.annot(
  pop.marker = 1e4,
  qtn.num = list(tr1 = 10, tr2 = 10),
  qtn.model = "A",
  qtn.dist = list(tr1 = "beta", tr2 = "beta"),
  qtn.shape1 = list(tr1 = 1, tr2 = 1),
  qtn.shape2 = list(tr1 = 1, tr2 = 1),
  qtn.ncp = list(tr1 = 0, tr2 = 0)
)
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2) # random genotype
# Generate phenotype simulation parameters
SP <- param.pheno(
  SP = SP,
  phe.model = list(
    tr1 = "T1 = A + E", # "T1" (Trait 1) consists of Additive effect and Residual effect
    tr2 = "T2 = A + E"  # "T2" (Trait 2) consists of Additive effect and Residual effect
  ),
  # phe.var = list(tr1 = 100, tr2 = 100),
  phe.h2A = list(tr1 = 0.3, tr2 = 0.3),
  phe.corA = matrix(c(1, 0.5, 0.5, 1), 2, 2) # Additive genetic correlation
)

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)

Generate phenotype with fixed effect and covariate and environmental random effect

back to top

SIMER supports adding Fixed effects, Covariates, and Environmental Random effects to a phenotype. Users should prepare a list of environmental factors setting. Fixed effects, Covariates , and Environmental Random effects are determined by effect, slope, and ratio respectively. A phenotype with Fixed effect, Covariate, and Environmental Random effect in single-trait simulation is displayed as follows:
If users want to output files, please see File output.

# Prepare environmental factor list
pop.env <- list(
  F1 = list( # fixed effect 1
    level = c("1", "2"),
    effect = list(tr1 = c(50, 30))
  ), 
  F2 = list( # fixed effect 2
    level = c("d1", "d2", "d3"),
    effect = list(tr1 = c(10, 20, 30))
  ),
  C1 = list( # covariate 1
    level = c(70, 80, 90),
    slope = list(tr1 = 1.5)
  ),
  R1 = list( # random effect 1
    level = c("l1", "l2", "l3"),
    ratio = list(tr1 = 0.1)
  )
)

# Generate genotype simulation parameters
SP <- param.annot(pop.marker = 1e4, qtn.num = list(tr1 = 10), qtn.model = "A")
# Generate annotation simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2) # random genotype
# Generate phenotype simulation parameters
SP <- param.pheno(
  SP = SP, 
  pop.env = pop.env,
  phe.model = list(tr1 = "T1 = A + F1 + F2 + C1 + R1 + E"), # "T1" (Trait 1) consists of Additive effect, F1, F2, C1, R1, and Residual effect
  # phe.var = list(tr1 = 100),
  phe.h2A = list(tr1 = 0.3)
)

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)

A phenotype with Fixed effect, Covariate, and Environmental Random effect in multiple-trait simulation is displayed as follows:
If users want to output files, please see File output.

# Prepare environmental factor list
pop.env <- list(
  F1 = list( # fixed effect 1
    level = c("1", "2"),
    effect = list(tr1 = c(50, 30), tr2 = c(50, 30))
  ), 
  F2 = list( # fixed effect 2
    level = c("d1", "d2", "d3"),
    effect = list(tr1 = c(10, 20, 30), tr2 = c(10, 20, 30))
  ),
  C1 = list( # covariate 1
    level = c(70, 80, 90),
    slope = list(tr1 = 1.5, tr2 = 1.5)
  ),
  R1 = list( # random effect 1
    level = c("l1", "l2", "l3"),
    ratio = list(tr1 = 0.1, tr2 = 0.1)
  )
)

# Generate genotype simulation parameters
SP <- param.annot(pop.marker = 1e4, qtn.num = list(tr1 = 10, tr2 = 10), qtn.model = "A")
# Generate annotation simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2) # random genotype
# Generate phenotype simulation parameters
SP <- param.pheno(
  SP = SP, 
  pop.env = pop.env,
  phe.model = list(
    tr1 = "T1 = A + F1 + F2 + C1 + R1 + E", # "T1" (Trait 1) consists of Additive effect, F1, F2, C1, R1, and Residual effect
    tr2 = "T2 = A + F1 + F2 + C1 + R1 + E"  # "T2" (Trait 1) consists of Additive effect, F1, F2, C1, R1, and Residual effect
  ),
  # phe.var = list(tr1 = 100, tr2 = 100),
  phe.h2A = list(tr1 = 0.3, tr2 = 0.3),
  phe.corA = matrix(c(1, 0.5, 0.5, 1), 2, 2) # Additive genetic correlation
)

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)

Generate phenotype using GxE model

back to top

In a "GxE" model, SIMER adds a Genetic-Environmental interaction effect to the phenotype. Users should prepare the Genetic QTN effect in the Annotation data and environmental factor by pop.env to generate a Genetic-Environmental Individual effect. An example of a Genetic-Environmental interaction in a single-trait simulation is displayed as follows:
If users want to output files, please see File output.

# Prepare environmental factor list
pop.env <- list(
  F1 = list( # fixed effect 1
    level = c("1", "2"),
    effect = list(tr1 = c(50, 30))
  ), 
  F2 = list( # fixed effect 2
    level = c("d1", "d2", "d3"),
    effect = list(tr1 = c(10, 20, 30))
  ),
  C1 = list( # covariate 1
    level = c(70, 80, 90),
    slope = list(tr1 = 1.5)
  ),
  R1 = list( # random effect 1
    level = c("l1", "l2", "l3"),
    ratio = list(tr1 = 0.1)
  )
)

# Generate annotation simulation parameters
SP <- param.annot(pop.marker = 1e4, qtn.num = list(tr1 = 10), qtn.model = "A") # Additive effect
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2) # random genotype
# Generate phenotype simulation parameters
SP <- param.pheno(
  SP = SP,
  pop.env = pop.env,
  phe.model = list(
    tr1 = "T1 = A + F1 + F2 + C1 + R1 + A:F1 + E" # "T1" (Trait 1) consists of Additive effect, F1, F2, C1, R1, Additive-F1 interaction effect, and Residual effect
  ),
  # phe.var = list(tr1 = 100),
  phe.h2A = list(tr1 = 0.3),
  phe.h2GxE = list(tr1 = list("A:F1" = 0.1))
)

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)

An example of Genetic-Environmental interaction in multiple-trait simulation is displayed as follows:
If users want to output files, please see File output.

# Prepare environmental factor list
pop.env <- list(
  F1 = list( # fixed effect 1
    level = c("1", "2"),
    effect = list(tr1 = c(50, 30), tr2 = c(50, 30))
  ), 
  F2 = list( # fixed effect 2
    level = c("d1", "d2", "d3"),
    effect = list(tr1 = c(10, 20, 30), tr2 = c(10, 20, 30))
  ),
  C1 = list( # covariate 1
    level = c(70, 80, 90),
    slope = list(tr1 = 1.5, tr2 = 1.5)
  ),
  R1 = list( # random effect 1
    level = c("l1", "l2", "l3"),
    ratio = list(tr1 = 0.1, tr2 = 0.1)
  )
)

# Generate annotation simulation parameters
SP <- param.annot(pop.marker = 1e4, qtn.num = list(tr1 = 10, tr2 = 10), qtn.model = "A") # Additive effect
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2) # random genotype
# Generate phenotype simulation parameters
SP <- param.pheno(
  SP = SP,
  pop.env = pop.env,
  phe.model = list(
    tr1 = "T1 = A + F1 + F2 + C1 + R1 + A:F1 + E", # "T1" (Trait 1) consists of Additive effect, F1, F2, C1, R1, Additive-F1 interaction effect, and Residual effect
    tr2 = "T2 = A + F1 + F2 + C1 + R1 + A:F1 + E"  # "T2" (Trait 2) consists of Additive effect, F1, F2, C1, R1, Additive-F1 interaction effect, and Residual effect
  ),
  # phe.var = list(tr1 = 100, tr2 = 100),
  phe.h2A = list(tr1 = 0.3, tr2 = 0.3),
  phe.h2GxE = list(tr1 = list("A:F1" = 0.1), tr2 = list("A:F1" = 0.1)),
  phe.corA = matrix(c(1, 0.5, 0.5, 1), 2, 2) # Additive genetic correlation
)

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)

Generate phenotype using ExE model

back to top

In an "ExE" model, SIMER adds Environmental-Environmental interaction effect to phenotype. Users should prepare environmental factor by pop.env for generating Environmental-Environmental Individual effect. An example of Environmental-Environmental interaction in single-trait simulation is displayed as follows:
If users want to output files, please see File output.

# Prepare environmental factor list
pop.env <- list(
  F1 = list( # fixed effect 1
    level = c("1", "2"),
    effect = list(tr1 = c(50, 30))
  ), 
  F2 = list( # fixed effect 2
    level = c("d1", "d2", "d3"),
    effect = list(tr1 = c(10, 20, 30))
  ),
  C1 = list( # covariate 1
    level = c(70, 80, 90),
    slope = list(tr1 = 1.5)
  ),
  R1 = list( # random effect 1
    level = c("l1", "l2", "l3"),
    ratio = list(tr1 = 0.1)
  )
)

# Generate annotation simulation parameters
SP <- param.annot(pop.marker = 1e4, qtn.num = list(tr1 = 10), qtn.model = "A") # Additive effect
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2) # random genotype
# Generate phenotype simulation parameters
SP <- param.pheno(
  SP = SP,
  pop.env = pop.env,
  phe.model = list(
    tr1 = "T1 = A + F1 + F2 + C1 + R1 + F1:R1 + E" # "T1" (Trait 1) consists of Additive effect, F1, F2, C1, R1, F1-R1 interaction effect, and Residual effect
  ),
  # phe.var = list(tr1 = 100),
  phe.h2A = list(tr1 = 0.3),
  phe.h2GxE = list(tr1 = list("F1:R1" = 0.1))
)

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)

An example of Environmental-Environmental interaction in multiple-trait simulation is displayed as follows:
If users want to output files, please see File output.

# Prepare environmental factor list
pop.env <- list(
  F1 = list( # fixed effect 1
    level = c("1", "2"),
    effect = list(tr1 = c(50, 30), tr2 = c(50, 30))
  ), 
  F2 = list( # fixed effect 2
    level = c("d1", "d2", "d3"),
    effect = list(tr1 = c(10, 20, 30), tr2 = c(10, 20, 30))
  ),
  C1 = list( # covariate 1
    level = c(70, 80, 90),
    slope = list(tr1 = 1.5, tr2 = 1.5)
  ),
  R1 = list( # random effect 1
    level = c("l1", "l2", "l3"),
    ratio = list(tr1 = 0.1, tr2 = 0.1)
  )
)

# Generate annotation simulation parameters
SP <- param.annot(pop.marker = 1e4, qtn.num = list(tr1 = 10, tr2 = 10), qtn.model = "A") # Additive effect
# Generate genotype simulation parameters
# SP <- param.geno(SP = SP, pop.geno = pop.geno)           # external genotype
SP <- param.geno(SP = SP, pop.ind = 1e2) # random genotype
# Generate phenotype simulation parameters
SP <- param.pheno(
  SP = SP,
  pop.env = pop.env,
  phe.model = list(
    tr1 = "T1 = A + F1 + F2 + C1 + R1 + F1:R1 + E", # "T1" (Trait 1) consists of Additive effect, F1, F2, C1, R1, F1:R1 interaction effect, and Residual effect
    tr2 = "T2 = A + F1 + F2 + C1 + R1 + F1:R1 + E"  # "T2" (Trait 2) consists of Additive effect, F1, F2, C1, R1, F1:R1 interaction effect, and Residual effect
  ),
  # phe.var = list(tr1 = 100, tr2 = 100),
  phe.h2A = list(tr1 = 0.3, tr2 = 0.3),
  phe.h2GxE = list(tr1 = list("F1:R1" = 0.1), tr2 = list("F1:R1" = 0.1)),
  phe.corA = matrix(c(1, 0.5, 0.5, 1), 2, 2) # Additive genetic correlation
)

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)

Generate phenotype controlled by varied QTN effect distribution

back to top

In the single-trait simulation, the trait can be controlled by varied QTN effect distribution. An example of the single-trait controlled by two-group QTNs is displayed as follows:
If users want to output files, please see File output.

# Generate annotation simulation parameters
SP <- param.annot(
  pop.marker = 1e4, 
  qtn.num = list(tr1 = c(2, 8)), # Group1: 2 QTNs; Group 2: 8 QTNs
  qtn.dist = list(tr1 = c("norm", "norm")),
  qtn.var = list(tr1 = c(1, 1)), # Group1: genetic variance of QTNs = 1; Group2: genetic variance of QTNs = 1
  qtn.model = "A"
)
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2) # random genotype
# Generate phenotype simulation parameters
SP <- param.pheno(
  SP = SP,
  phe.model = list(tr1 = "T1 = A + E"), # "T1" (Trait 1) consists of Additive effect and Residual effect
  # phe.var = list(tr1 = 100),
  phe.h2A = list(tr1 = 0.3)
)

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)

Population Simulation of Multiple-Generation with Genotype and Phenotype

back to top

SIMER imitates the reproductive process of organisms to generate a Multiple-Generation population. The genotype data and phenotype data of the population are screened by single-trait selection or multiple-trait selection, and then those data are amplified by species-specific reproduction.

Gallery of population simulation parameters

back to top

selects, main function of Selection:

reproduces, main function of Reproduction:

Individual selection for a single trait

back to top

Individual selection is a selection method based on the phenotype of individual traits, which is also known as mixed selection or collective selection. This selection method is simple and easy to use for traits with high heritability.
If users want to output files, please see File output.

# Generate annotation simulation parameters
SP <- param.annot(pop.marker = 1e4)
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2)
# Generate phenotype simulation parameters
SP <- param.pheno(SP = SP, phe.h2A = list(tr1 = 0.3))
# Generate selection parameters
SP <- param.sel(SP = SP, sel.single = "ind")

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)
# Run selection
SP <- selects(SP)

Family selection for a single trait

back to top

Family selection is a selection method by family based on the average of the family. This selection method is used for traits with low heritability.
If users want to output files, please see File output.

# Generate annotation simulation parameters
SP <- param.annot(pop.marker = 1e4)
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2)
# Generate phenotype simulation parameters
SP <- param.pheno(SP = SP, phe.h2A = list(tr1 = 0.3))
# Generate selection parameters
SP <- param.sel(SP = SP, sel.single = "fam")

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)
# Run selection
SP <- selects(SP)

Within-family selection for a single trait

back to top

Within-family selection is a selection method based on the deviation of individual phenotype and family mean value in each family. This selection method is used for traits with low heritability and small families.
If users want to output files, please see File output.

# Generate annotation simulation parameters
SP <- param.annot(pop.marker = 1e4)
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2)
# Generate phenotype simulation parameters
SP <- param.pheno(SP = SP, phe.h2A = list(tr1 = 0.3))
# Generate selection parameters
SP <- param.sel(SP = SP, sel.single = "infam")

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)
# Run selection
SP <- selects(SP)

Combined selection for a single trait

back to top

Combined selection is a selection method based on weighed combination of the deviation of individual phenotype and family mean value.
If users want to output files, please see File output.

# Generate annotation simulation parameters
SP <- param.annot(pop.marker = 1e4)
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2)
# Generate phenotype simulation parameters
SP <- param.pheno(SP = SP, phe.h2A = list(tr1 = 0.3))
# Generate selection parameters
SP <- param.sel(SP = SP, sel.single = "comb")

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)
# Run selection
SP <- selects(SP)

Tandem selection for multiple traits

back to top

Tandem selection is a method for sequentially selecting a plurality of target traits one by one. The index of the selected trait is index.tdm and this parameter should not be controlled by Users.
If users want to output files, please see File output.

# Generate genotype simulation parameters
SP <- param.annot(pop.marker = 1e4, qtn.num = list(tr1 = 10, tr2 = 10))
# Generate annotation simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2)
# Generate phenotype simulation parameters
SP <- param.pheno(
  SP = SP, 
  # phe.var = list(tr1 = 100, tr2 = 100),
  phe.model = list(
    tr1 = "T1 = A + E",
    tr2 = "T2 = A + E"
  )
)
# Generate selection parameters
SP <- param.sel(SP = SP, sel.multi = "tdm")

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)
# Run selection
SP <- selects(SP)

Independent culling selection for multiple traits

back to top

Set a minimum selection criterion for each target trait. Then a Independent culling selection will eliminate this individual when the candidate's performance on any trait is lower than the corresponding criteria.
If users want to output files, please see File output.

# Generate genotype simulation parameters
SP <- param.annot(pop.marker = 1e4, qtn.num = list(tr1 = 10, tr2 = 10))
# Generate annotation simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2)
# Generate phenotype simulation parameters
SP <- param.pheno(
  SP = SP, 
  # phe.var = list(tr1 = 100, tr2 = 100),
  phe.model = list(
    tr1 = "T1 = A + E",
    tr2 = "T2 = A + E"
  )
)
# Generate selection parameters
SP <- param.sel(SP = SP, sel.multi = "indcul")

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)
# Run selection
SP <- selects(SP)

Index selection for multiple traits

back to top

Index selection is a comprehensive selection that will consider several traits based on their respective heritabilities, phenotypic variances, economic weights, corresponding genetic correlations, and phenotypes. Then, SIMER calculates the index value of each trait, eliminates it, or selects it according to its level. Users can set the weight of each trait at index.wt.
If users want to output files, please see File output.

# Generate genotype simulation parameters
SP <- param.annot(pop.marker = 1e4, qtn.num = list(tr1 = 10, tr2 = 10))
# Generate annotation simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2)
# Generate phenotype simulation parameters
SP <- param.pheno(
  SP = SP, 
  # phe.var = list(tr1 = 100, tr2 = 100),
  phe.model = list(
    tr1 = "T1 = A + E",
    tr2 = "T2 = A + E"
  )
)
# Generate selection parameters
SP <- param.sel(SP = SP, sel.multi = "index")

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)
# Run selection
SP <- selects(SP)

Clone for plants

back to top

Clone is a sexual reproduction method that does not involve germ cells and does not require a process of fertilization, but directly forms a new individual's reproductive mode from a part of the mother. Sex of offspring will be 0 in the clone.
If users want to output files, please see File output.

# Generate annotation simulation parameters
SP <- param.annot(pop.marker = 1e4)
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2)
# Generate phenotype simulation parameters
SP <- param.pheno(SP = SP, phe.h2A = list(tr1 = 0.3))
# Generate selection parameters
SP <- param.sel(SP = SP, sel.single = "ind")
# Generate reproduction parameters
SP <- param.reprod(SP = SP, reprod.way = "clone")

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)
# Run selection
SP <- selects(SP)
# Run reproduction
SP <- reproduces(SP)

Doubled haploid for plants

back to top

Doubled haploid is a reproduction method for breeding workers to obtain haploid plants. It induces a doubling of the number of chromosomes and restores the number of chromosomes in normal plants. Sex of offspring will be 0 in dh.
If users want to output files, please see File output.

# Generate annotation simulation parameters
SP <- param.annot(pop.marker = 1e4)
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2)
# Generate phenotype simulation parameters
SP <- param.pheno(SP = SP, phe.h2A = list(tr1 = 0.3))
# Generate selection parameters
SP <- param.sel(SP = SP, sel.single = "ind")
# Generate reproduction parameters
SP <- param.reprod(SP = SP, reprod.way = "dh")

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)
# Run selection
SP <- selects(SP)
# Run reproduction
SP <- reproduces(SP)

Self-pollination for plants and micro-organisms

back to top

Self-pollination refers to the combination of male and female gametes from the same individual or between individuals from the same clonal breeding line. Sex of offspring will be 0 in selfpol.
If users want to output files, please see File output.

# Generate annotation simulation parameters
SP <- param.annot(pop.marker = 1e4)
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2)
# Generate phenotype simulation parameters
SP <- param.pheno(SP = SP, phe.h2A = list(tr1 = 0.3))
# Generate selection parameters
SP <- param.sel(SP = SP, sel.single = "ind")
# Generate reproduction parameters
SP <- param.reprod(SP = SP, reprod.way = "selfpol")

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)
# Run selection
SP <- selects(SP)
# Run reproduction
SP <- reproduces(SP)

Random mating for plants and animals

back to top

In random mating, any female or male individual has the same probability to mate with any member of opposite sex in a sexually reproducing organism. Sex of offspring in random mating is controlled by sex.ratio in randmate.
If users want to output files, please see File output.

# Generate annotation simulation parameters
SP <- param.annot(pop.marker = 1e4)
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2)
# Generate phenotype simulation parameters
SP <- param.pheno(SP = SP, phe.h2A = list(tr1 = 0.3))
# Generate selection parameters
SP <- param.sel(SP = SP, sel.single = "ind")
# Generate reproduction parameters
SP <- param.reprod(SP = SP, reprod.way = "randmate")

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)
# Run selection
SP <- selects(SP)
# Run reproduction
SP <- reproduces(SP)

Random mating excluding self-pollination for animals

back to top

In random mating excluding self-pollination, an individual cannot mate with itself. Sex of offspring in random mating is controlled by sex.ratio in randexself.
If users want to output files, please see File output.

# Generate annotation simulation parameters
SP <- param.annot(pop.marker = 1e4)
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2)
# Generate phenotype simulation parameters
SP <- param.pheno(SP = SP, phe.h2A = list(tr1 = 0.3))
# Generate selection parameters
SP <- param.sel(SP = SP, sel.single = "ind")
# Generate reproduction parameters
SP <- param.reprod(SP = SP, reprod.way = "randexself")

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)
# Run selection
SP <- selects(SP)
# Run reproduction
SP <- reproduces(SP)

Assortative mating for plants and animals

back to top

In assortative mating, mated pairs are of the same phenotype more often than would occur by chance. Sex of offspring in assortative mating is controlled by sex.ratio in assort.
If users want to output files, please see File output.

# Generate annotation simulation parameters
SP <- param.annot(pop.marker = 1e4)
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2)
# Generate phenotype simulation parameters
SP <- param.pheno(SP = SP, phe.h2A = list(tr1 = 0.3))
# Generate selection parameters
SP <- param.sel(SP = SP, sel.single = "ind")
# Generate reproduction parameters
SP <- param.reprod(SP = SP, reprod.way = "assort")

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)
# Run selection
SP <- selects(SP)
# Run reproduction
SP <- reproduces(SP)

Disassortative mating for plants and animals

back to top

In disassortative mating, mated pairs are of the same phenotype less often than would occur by chance. Sex of offspring in disassortative mating is controlled by sex.ratio in disassort.
If users want to output files, please see File output.

# Generate annotation simulation parameters
SP <- param.annot(pop.marker = 1e4)
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2)
# Generate phenotype simulation parameters
SP <- param.pheno(SP = SP, phe.h2A = list(tr1 = 0.3))
# Generate selection parameters
SP <- param.sel(SP = SP, sel.single = "ind")
# Generate reproduction parameters
SP <- param.reprod(SP = SP, reprod.way = "disassort")

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)
# Run selection
SP <- selects(SP)
# Run reproduction
SP <- reproduces(SP)

Two-way cross for animals

back to top

The Two-way cross method needs to use sex to distinguish two different breeds, in which the first breed is sire and the second breed is dam.
If users want to output files, please see File output.

# Generate annotation simulation parameters
SP <- param.annot(pop.marker = 1e4)
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2)
# Generate phenotype simulation parameters
SP <- param.pheno(SP = SP, phe.h2A = list(tr1 = 0.3))
# Generate selection parameters
SP <- param.sel(SP = SP, sel.single = "ind")
# Generate reproduction parameters
SP <- param.reprod(SP = SP, reprod.way = "2waycro")

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)
# Two different breeds are cut by sex
SP$pheno$pop$gen1$sex <- rep(c(1, 2), c(50, 50))
# Run selection
SP <- selects(SP)
# Run reproduction
SP <- reproduces(SP)

Three-way cross for animals

back to top

The Three-way cross method needs to use sex to distinguish three different breeds, in which the first breed is sire and the second breed is dam in the first two-way cross, and the third breed is terminal sire.
If users want to output files, please see File output.

# Generate annotation simulation parameters
SP <- param.annot(pop.marker = 1e4)
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2)
# Generate phenotype simulation parameters
SP <- param.pheno(SP = SP, phe.h2A = list(tr1 = 0.3))
# Generate selection parameters
SP <- param.sel(SP = SP, sel.single = "ind")
# Generate reproduction parameters
SP <- param.reprod(SP = SP, reprod.way = "3waycro")

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)
# Three different breeds are cut by sex
SP$pheno$pop$gen1$sex <- rep(c(1, 2, 1), c(30, 30, 40))
# Run selection
SP <- selects(SP)
# Run reproduction
SP <- reproduces(SP)

Four-way cross for animals

back to top

The Four-way cross method needs to use sex to distinguish four different breeds, in which the first breed is sire and the second breed is dam in the first two-way cross, the third breed is sire and the fourth breed is dam in the second two-way cross.
If users want to output files, please see File output.

# Generate annotation simulation parameters
SP <- param.annot(pop.marker = 1e4)
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2)
# Generate phenotype simulation parameters
SP <- param.pheno(SP = SP, phe.h2A = list(tr1 = 0.3))
# Generate selection parameters
SP <- param.sel(SP = SP, sel.single = "ind")
# Generate reproduction parameters
SP <- param.reprod(SP = SP, reprod.way = "4waycro")

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)
# Three different breeds are cut by sex
SP$pheno$pop$gen1$sex <- rep(c(1, 2, 1, 2), c(25, 25, 25, 25))
# Run selection
SP <- selects(SP)
# Run reproduction
SP <- reproduces(SP)

Back cross for animal

back to top

The Back cross method needs to use sex to distinguish two different breeds, in which the first breed is always sire in each generation and the second breed is dam in the first two-way cross.
If users want to output files, please see File output.

# Generate annotation simulation parameters
SP <- param.annot(pop.marker = 1e4)
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2)
# Generate phenotype simulation parameters
SP <- param.pheno(SP = SP, phe.h2A = list(tr1 = 0.3))
# Generate selection parameters
SP <- param.sel(SP = SP, sel.single = "ind")
# Generate reproduction parameters
SP <- param.reprod(SP = SP, reprod.way = "backcro")

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)
# Two different breeds are cut by sex
SP$pheno$pop$gen1$sex <- rep(c(1, 2), c(50, 50))
# Run selection
SP <- selects(SP)
# Run reproduction
SP <- reproduces(SP)

User-designed pedigree mating for plants and animals

back to top

User-designed pedigree mating needs a specific user-designed pedigree to control the mating process. The first column is sample id, the second column is paternal id, and the third column is maternal id. Please make sure that paternal id and maternal id can match the genotype data.
If users want to output files, please see File output.

# Generate annotation simulation parameters
SP <- param.annot(pop.marker = 1e4)
# Generate genotype simulation parameters
SP <- param.geno(SP = SP, pop.ind = 1e2)
# Generate phenotype simulation parameters
SP <- param.pheno(SP = SP, phe.h2A = list(tr1 = 0.3))
# Generate reproduction parameters
SP <- param.reprod(SP = SP, reprod.way = "userped")

# Run annotation simulation
SP <- annotation(SP)
# Run genotype simulation
SP <- genotype(SP)
# Run phenotype simulation
SP <- phenotype(SP)
# Run reproduction
SP <- reproduces(SP)

AN EASY WAY TO GENERATE A POPULATION

back to top

The above methods are to generate populations step by step, which are easy to understand. Actually, SIMER can generate a population directly in a MORE CONVENIENT way.
If users want to output files, please see File output.

# Generate all simulation parameters
SP <- param.simer(qtn.num = list(tr1 = 10), pop.marker = 1e4, pop.ind = 1e2, sel.single = "comb", reprod.way = "randmate")

# Run Simer
SP <- simer(SP)

Breeding Program Design

back to top

After generating a population, further work can be done. Breeders wish to evaluate their Breeding Program Design. To save money and time, SIMER can assist breeders to evaluate their Breeding Program Design by simulation.

Gallery of breeding program design parameters

back to top

simer.Data.Json, main function of Breeding Program Design:

Preparation of a breeding program design

back to top

Breeding program design should be stored on a JSON file.

plan1.json

genotype: the absolute path or relative path to JSON file of genotype data
pedigree: the filename with absolute path or relative path to JSON file of pedigree data
selection_index: the economic weight of phenotype for each trait
threads: the threads number used in multiple threads computation
genetic_progress: the genetic progress of a breeding plan
breeding_value_index: the economic weight of breeding value for each trait
auto_optimization: optimizing EBV estimated model and selection index automatically
quality_control_plan: the quality control plan for genotype, pedigree, and phenotype

genotype_quality_control: the quality control plan for genotype

filter: the 'filter' (individual) condition for genotyped individual
filter_geno: the genotype missing rate filter
filter_mind the sample missing rate filter
filter_maf the Minor Allele Frequency filter
filter_hwe the Hardy-Weinberg Equilibrium filter

pedigree_quality_control: the quality control plan for pedigree

standard_ID: whether ID is standard 15-digit ID
candidate_sire_file: the filename of candidate sire
candidate_dam_file: the filename of candidate dam
exclude_threshold: if the number of base error is more than this threshold, this individual will be excluded
assign_threshold: if the number of base error is less than this threshold, this parent will be assigned to this individual

phenotype_quality_control: the quality control plan for phenotype

job_name: the name of phenotype quality control job
sample_info: the filename with absolute path or relative path to JSON file of phenotype data
repeated_records: whether phenotype data contains repeated records
multi_trait: whether phenotype data contains multiple traits
filter: the 'filter' (individual) condition for phenotyped individual
select: the 'select' (trait) condition for phenotyped individual
arrange: the 'arrange' (order) condition for phenotyped individual
job_traits: the trait need quality control and its definition and range

breeding_plan: the genetic evaluation plan

job_name: the name of phenotype quality control job
sample_info: the filename with absolute path or relative path to JSON file of phenotype data
repeated_records: whether phenotype data contains repeated records
multi_trait: whether phenotype data contains multiple traits
vc_vars: the filename of variance component data
vc_covars: the filename of covariance component data
random_ratio: the least random effect ratio to phenotype variance
job_traits: the trait need analysis and its covariate, fixed effect, and random effect

{
    "genotype": "../02plinkb",
    "pedigree": "../05others/pedigree.txt",
    "selection_index": "100 - 0.2 * T1 + 0.8 * T2",
    "threads": 16,
    "genetic_progress": [],
    "breeding_value_index": "-0.2 * T1 + 0.8 * T2",
    "auto_optimization": true,
    "quality_control_plan": {
        "genotype_quality_control":{
            "filter": "F1 == 'Male'",
            "filter_geno": 0.1,
            "filter_mind": 0.1,
            "filter_maf": 0.05,
            "filter_hwe": 0.001
        },
        "pedigree_quality_control":{
            "standard_ID": false,
            "candidate_sire_file": [],
            "candidate_dam_file": [],
            "exclude_threshold": 0.1, 
            "assign_threshold": 0.05
        },
        "phenotype_quality_control":[
            {
                "job_name": "Data_Quality_Control_Demo",
                "sample_info": "../05others/phenotype.txt",
                "repeated_records": false,
                "multi_trait": true,
                "filter": "F1 == 'Male'",
                "job_traits": [
                    {
                        "traits": "T1",
                        "definition": "T1",
                        "range": []
                    },
                    {
                        "traits": "T2",
                        "definition": "T2",
                        "range": []
                    }
                ]
            }
        ]
    },
    "breeding_plan":[
        {
            "job_name": "EBV_Model_Demo",
            "sample_info": "../05others/phenotype.txt",
            "repeated_records": false,
            "multi_trait": true,
            "vc_vars": [],
            "vc_covars": [],
            "random_ratio": 0.05,
            "job_traits": [
                {
                    "traits": "T1",
                    "covariates": [],
                    "fixed_effects": ["F1", "F2"],
                    "random_effects": ["R1"]
                },
                {
                    "traits": "T2",
                    "covariates": [],
                    "fixed_effects": ["F1", "F2"],
                    "random_effects": ["R1"]
                }
            ]
        }
    ]
}

Evaluation of a breeding program design

back to top

To evaluate the breeding program design, SIMER completes the following three tasks:
(1) Data quality control for genotype, pedigree, and phenotype
(2) Model optimization (i.e., the most suitable covariate, fixed effect, and random effect)
(3) Construction of Selection Index and calculation of Genetic Progress

# Get JSON file
jsonFile <- system.file("extdata", "04breeding_plan", "plan1.json", package = "simer")

# It needs 'plink' and 'hiblup' software
jsonList <- simer.Data.Json(jsonFile = jsonFile)

Global Options

back to top

Users can use global parameters to control the population properties , the number of threads used for simulation, and the output of simulation data.

Gallery of global parameters

back to top

simer, main function of simulation:

Counts of total population size

back to top

Users can calculate the number of individuals per generation using IndPerGen directly.

pop <- generate.pop(pop.ind = 100)
count.ind <- IndPerGen(pop = pop, pop.gen = 2, ps = c(0.8, 0.8), reprod.way = "randmate", sex.rate = 0.5, prog = 2)

Multi-thread computation

back to top

SIMER runs on multiple threads. Users can easily change the number of threads used for simulation by the following:

# Generate all simulation parameters
SP <- param.simer(out = "simer", ncpus = 2)

# Run Simer
SP <- simer(SP)

Multi-population simulation

back to top

Simulation of multiple populations can be realized by for by using R software.

# Replication times
rep <- 2

# Result list
SPs <- rep(list(NULL), rep)

for (i in 1:rep) {
  # Generate all simulation parameters
  SP <- param.simer(replication = i, seed.sim = i, out = "simer")

  # Run Simer
  SPs[[i]] <- simer(SP)
}

File output

back to top

SIMER will not output files by default. A series of files with the prefix out will output when specifying outpath.

### 01 Numeric Format ###
# Generate all simulation parameters
SP <- param.simer(
  # SP = SP, # uncomment it when users already have a 'SP'
  out = "simer",
  outpath = getwd(),
  out.format = "numeric"
 )

# Run Simer
SP <- simer(SP)

### 02 PLINK Binary Format ###
# Generate all simulation parameters
SP <- param.simer(
  # SP = SP, # uncomment it when users already have a 'SP'
  out = "simer",
  outpath = getwd(),
  out.format = "plink"
)

# Run Simer
SP <- simer(SP)

Generation-selective output

back to top

Output of genotype and phenotype can be generation-selective using out.geno.gen and out.pheno.gen.

# Generate all simulation parameters
SP <- param.simer(out = "simer", outpath = getwd(), pop.gen = 2, out.geno.gen = 1:2, out.pheno.gen = 1:2)

# Run Simer
SP <- simer(SP)

Output

back to top

SIMER outputs data including annotation data, genotype data, and phenotype data in the following two format.
Numeric format:
simer.geno.ind contains indice of genotyped individuals;
simer.geno.desc and simer.geno.bin contain genotype matrix of all individuals;
simer.map contains input map with block information and recombination information;
simer.ped contains pedigree of individuals;
simer.phe contains phenotype of individuals.
PLINK Binary format:
simer.bim contains marker information of genotype data;
simer.bed contains genotype data in binary format;
simer.fam contains sample information of genotype data;
simer.ped contains pedigree of individuals;
simer.phe contains phenotype of individuals.

Annotation data

back to top

Annotation data contains SNP name, Chromosome name, Base Position, ALT, REF, and the QTN genetic effect. Note that only markers selected as QTNs have values.

# Generate all simulation parameters
SP <- param.simer(out = "simer")

# Run Simer
SP <- simer(SP)

# Show annotation data
head(SP$map$pop.map)
  SNP Chrom     BP ALT REF QTN1_A
1  M1     1 130693   C   A     NA
2  M2     1 168793   G   A     NA
3  M3     1 286553   A   T     NA
4  M4     1 306913   C   G     NA
5  M5     1 350926   T   A     NA
6  M6     1 355889   A   C     NA

Genotype data

back to top

Genotype data is stored in big.matrix format.

# Generate all simulation parameters
SP <- param.simer(out = "simer")

# Run Simer
SP <- simer(SP)

# Show genotype data
print(SP$geno$pop.geno)
$gen1
An object of class "big.matrix"
Slot "address":
<pointer: 0x00000000176f09e0>


$gen2
An object of class "big.matrix"
Slot "address":
<pointer: 0x00000000176ef940>

print(SP$geno$pop.geno$gen1[1:6, 1:6])
     [,1] [,2] [,3] [,4] [,5] [,6]
[1,]    0    2    0    1    0    2
[2,]    1    1    1    1    0    0
[3,]    0    1    2    2    1    0
[4,]    2    0    1    1    1    0
[5,]    2    1    0    1    2    1
[6,]    1    2    1    1    1    2

Phenotype data

back to top

Phenotype data contains sample ID, generation index, family index, within-family index, sire, dam, sex, phenotype, TBV, TGV, and other effects.

# Generate all simulation parameters
SP <- param.simer(out = "simer")

# Run Simer
SP <- simer(SP)

# Show phenotype data
head(SP$pheno$pop$gen1)
  index gen fam infam sir dam sex          T1     T1_TBV     T1_TGV   T1_A_eff    T1_E_eff
1     1   1   1     1   0   0   1  -0.4934935 -1.3507888 -1.3507888 -1.3507888   0.8572953
2     2   1   2     2   0   0   1   7.7710404 -1.6756353 -1.6756353 -1.6756353   9.4466757
3     3   1   3     3   0   0   1  -4.6567338 -2.2608387 -2.2608387 -2.2608387  -2.3958951
4     4   1   4     4   0   0   1  -5.9064589 -1.7394139 -1.7394139 -1.7394139  -4.1670450
5     5   1   5     5   0   0   1 -16.7438931 -2.8000846 -2.8000846 -2.8000846 -13.9438085
6     6   1   6     6   0   0   1   6.0043912  0.3413561  0.3413561  0.3413561   5.6630351

Citation

back to top

For SIMER:
Hope it will be coming soon!

For ADI model:
Kao, Chenhung, et al. "Modeling Epistasis of Quantitative Trait Loci Using Cockerham's Model." Genetics 160.3 (2002): 1243-1261.

For build.cov:
B. D. Ripley "Stochastic Simulation." Wiley-Interscience (1987): Page 98.

FAQ and Hints

back to top

:sos: Question1: Failing to install "devtools":

ERROR: configuration failed for package ‘git2r’

removing ‘/Users/acer/R/3.4/library/git2r’

ERROR: dependency ‘git2r’ is not available for package ‘devtools’

removing ‘/Users/acer/R/3.4/library/devtools’

:yum: Answer: Please try following codes in terminal:

apt-get install libssl-dev/unstable

:sos: Question2: When installing packages from Github with "devtools", an error occurred:

Error in curl::curl_fetch_disk(url, x$path, handle = handle): Problem with the SSL CA cert (path? access rights?)

:yum: Answer: Please try following codes and then try agian.

library(httr)
set_config(config(ssl_verifypeer = 0L))

Questions, suggestions, and bug reports are welcome and appreciated.:arrow_right:

back to top.