cash-addBIsnps: BC and FS method: Add BI SNPs to existing linkage groups

In tpbilton/GUSMap: Genotyping Uncertainty with Sequencing Data and Linkage Mapping (GUSMap)

BC and FS method: Add BI SNPs to existing linkage groups

Description

Method for adding both-informative (BI) SNPs to paternal and maternal linkage groups and to combine maternal and paternal linkage groups.

Usage

BCobj$addBIsnps(LODthres = 10, nComp = 30)
FSobj$addBIsnps(LODthres = 10, nComp = 30)

Arguments

LODthres	A positive numeric value specifying the LOD threshold used to add SNPs to the linkage groups.
nComp	A positive integer value specifying how many SNPs in the linkage group to compute the average LOD score with the unmapped SNP.

Details

Each both-informative (BI) SNP is mapped to the maternal and paternal linkage groups using the following algorithm:

1. For each MI (or PI) linkage group, compute the average LOD score between the unmapped BI SNP and a specified number (nComp) of SNPs in the MI (or PI) linkage group that have the highest LOD score with the unmapped BI SNP.

2. Map the unmapped BI SNP to linkage group with the highest average LOD score if:

   • The largest average LOD score is greater than the LOD threshold (LODthres).

   • All the other average LOD scores are less than the LOD threshold.

The performance of the linkage group algorithm depends on the value of LODthres and nComp. The user is advised to experiment with different values and examine the matrix of recombination fractions (using the $plotLG function).

BC object: Each BI SNP is mapped to a MI linkage group and then independently to a PI linkage group using the above algorithm. The linkage groups produced from this function are referred to as the "pseudo-testcross linkage groups with BI SNPs" and are stored separately to the "pseudo-testcross linkage groups" produced by $createLG function. This means that pseudo-testcross linkage groups with BI SNPs can be produced at any time from the pseudo-testcross linkage groups, even after major edits to the pseudo-testcross linkage groups with BI SNPs have been made.

FS object: Each BI SNP is mapped to a MI linkage group and then independently to a PI linkage group using the above algorithm. The MI and PI linkage groups are then merged together using the following algorithm:

1. Compute the number of BI SNPs that mapped to the same pair of MI and PI linkage groups.

2. Merge the pair of MI and PI linkage groups that has the largest number of BI SNPs mapped to them, provided that both linkage groups have not already been merged.

3. Contiune step 2 until either every MI linkage group has been merged with a PI linkage group, or every PI linkage group has been merged with a MI linkage group, or until no more MI or PI linkage groups can can be merged to another non-merged linkage group.

4. Merge the remaining (unmerged) MI or PI linkage groups to the combined linkage group with the most BI SNPs in common (provided there are any).

5. Contiune step 4 until all linkage groups have been merged or until no more linkage groups can be merged.

The linkage groups produced are referred to as the "combined linkage groups" and are stored separately to the "pseudo-testcross linkage groups" produced by $createLG function. This means that combined linkage groups can be produced at any time from the pseudo-testcross linkage groups, even after major edits to the combined linkage groups have been made.

Author(s)

Timothy P. Bilton

See Also

BC, FS

Examples

## Simulate some sequencing data
set.seed(6745)
config <- list(list(sample(c(1,2,4), size=30, replace=TRUE)))
F1data <- simFS(0.01, config=config, meanDepth=10, nInd = 50)
## Compute 2-point recombination fractions
F1data$rf_2pt()
## create paternal and maternal linkage groups
F1data$createLG()

## Add the BI SNPs
F1data$addBIsnps()

tpbilton/GUSMap documentation built on Feb. 22, 2025, 12:27 p.m.