R package:blupADC-Feature 6

Overview

🙉 In the application of animal and plant breeding, the key step is the construction of kinship matrix. Package: blupADC provides cal_kinship function which can construct various type of relationship matrix directly, including additive relationship matrix(pedigree, genomic and single-step), and dominance relationship matrix(pedigree, genomic and single-step), and the inverse matrix of these kinship matrix.

In the construction of single-step relationship matrix, users can select Metafounder algorithm or APY algorithm. In terms of the construction of dominance relationship matrix, users can select different coding manners for genomic dominance relationship matrix, gene dropping algorithm for pedigree dominance relationship matrix. In addition, cal_kinship can calculate several types of inbreeding coefficients (pedigree, genomic, and single-step) .

Note: In the construction of genomic and single-step relationship matrix, people need to provide genotype data. These parameters are the same as in geno_format function((see more details).

Example

😾 Pedigree-based kinship matrix

library(blupADC)
kinship_result=cal_kinship(
                input_pedigree=example_ped3,          #provided hapmap data object
                kinship_type=c("P_A"),      #type of  kinship matrix
                inbred_type=c("Pedigree"),      #type of inbreeding coefficients
                return_result=TRUE)               #return result              

🤡 Genomic-based kinship matrix

library(blupADC)
kinship_result=cal_kinship(
                input_data_hmp=example_data_hmp,          #provided hapmap data object
                kinship_type=c("G_A","G_D"),      #type of  kinship matrix
                dominance_type=c("genotypic"),    #type of dominance effect
                inbred_type=c("Homozygous"),      #type of inbreeding coefficients
                return_result=TRUE)               #return result              

💨Single-step based kinship matrix

library(blupADC)
kinship_result=cal_kinship(
                input_data_hmp=example_data_hmp,          #provided hapmap data object
                input_pedigree=example_ped3,
                kinship_type=c("H_A"),      #type of  kinship matrix
                inbred_type=c("H_diag"),      #type of inbreeding coefficients
                return_result=TRUE)               #return result              

💨Single-step based kinship matrix(via bigmemory method)

library(blupADC)
phased_kinship_result=cal_kinship(
                input_data_hmp=example_data_hmp,          #provided hapmap data object
                input_pedigree=example_ped3,
                bigmemory_cal=TRUE,
                bigmemory_data_path=getwd(),
                bigmemory_data_name="blupADC",
                kinship_type=c("H_A"),      #type of  kinship matrix
                inbred_type=c("H_diag"),      #type of inbreeding coefficients
                return_result=TRUE)               #return result                

Parameter

💪Basic

• 1：kinship_type

  Type of kinship matrix, character class. User can select multiple types simultaneously, including:

  • “G_A” : genomic additive kinship matrix
  • “G_Ainv” :inverse of genomic additive kinship matrix
  • “G_D” :genomic dominance kinship matrix
  • “G_Dinv” :inverse of genomic dominance kinship matrix
  • “P_A” :pedigree additive kinship matrix
  • “P_Ainv” :inverse of pedigree additive kinship matrix
  • “P_D” : pedigree dominance kinship matrix
  • “P_Dinv” :inverse of pedigree dominance kinship matrix
  • “H_A” :single-step additive kinship matrix
  • “H_Ainv” :inverse of single-step additive kinship matrix

Note: In the construction of pedigree and single-step relationship matrix, user need to provide pedigree data. In the construction of genomic and single-step relationship matrix, user need to provide genotype data.

• 2：dominance_type

  Type of dominance effect in the construction of dominance relationship matrix , character class.

  • “genotypic” : coded by $0-2pq$, $1-2pq$, and $0-2pq$ for $AA$, $Aa$, and $aa$, respectively.
  • “classical” : coded by $-2q^2$, $2pq$ , and $-2p^2$ for $AA$, $Aa$, and $aa$, respectively.

More details about these two types dominance effects could be seen in this reference ：On the Additive and Dominant Variance and Covariance of Individuals Within the Genomic Selection Scope

• 3：inbred_type

  Type of inbreeding coefficients, character class.

  • “Homozygous” : proportion of homozygous sites

  • “G_Diag” : diagonal of genomic additive relationship matrix minus 1

  • “H_diag” :diagonal of single-step additive relationship matrix minus 1

  • “Pedigree” :diagonal of pedigree additive relationship matrix minus 1

• 4：input_pedigree

User-provided pedigree data ，data.frame or matrix class. (see more details about the format of pedigree data)

• 5：IND_rename

Whether genotype individuals need to be renamed according to the provided pedigree, logical class. Default is FALSE.

• 6:bigmemory_cal

Whether using bigmemory method to calculate. logical class. Default is FALSE.

• 7:bigmemory_data_path

The file path bigmemory data . character class.

• 8:bigmemory_data_name

The file name bigmemory data . character class.

• 9：output_matrix_type

  Type of output kinship matrix type, character class. Default is “col_all”.

  • “col_all” : n*n format
  • “col_three” : 3 columns format. The first and the second column are the name of individuals, the third column is the relationship coefficients 。

• 10：output_matrix_path

File path of output relationship matrix, character class.

• 11：output_matrix_name

File name of output relationship matrix, character class.

🦶Advanced

• 12：cpu_cores

Number of cpu in calculating, numeric class. Default is 1.

• 13：kinship_base

Whether take $p=q=0.5$ in the construction of relationship matrix, logical class. Default is FALSE.

• 14：kinship_trace

Whether take the trace of kinship matrix to scale relationship matrix, logical class. Default is FALSE.

• 15：Metafounder_algorithm

Whether take the metafounder algorithm to construct single-step relationship matrix, logical class. Default is FALSE.

• 16：APY_algorithm

Whether take the APY algorithm to construct inverse relationship matrix, logical class. Default is FALSE.

• 17：APY_eigen_threshold

Threshold of variation explained by eigenvalues , numeric class. Default is 0.95.

• 18：APY_n_core

Number of core animals , numeric class. Default is NULL.

• 19：SSBLUP_omega

The value of omega in the construction of single-step additive relationship matrix, numeric class. Default is 0.05.

• 20：gene_dropping

Whether take the gene dropping algorithm to construct pedigree dominance relationship matrix, logical class. Default is FALSE.

• 21：gene_dropping_iteration

The number of iteration for gene dropping algorithm , numeric class. Default is 1000.