Plot LD vs distance using an asymptotic regression model.
Usage
LD_decay_plot(
data,
map,
max.pair = 10000,
max.loci = NULL,
position = "bp",
r2_threshold = 0.2,
xlim_range = c(0, 20),
show_vline = TRUE,
show_hline = TRUE,
seed = 123
)Arguments
- data
Genotype matrix (markers as columns, individuals as rows). Rownames have the marker names that matcho those of the marker column in the
map.- map
Data frame with columns 'chrom', 'marker', and 'position'
- max.pair
Maximum number of r² pairs for the model (default: 1e4)
- max.loci
Maximum number of markers to use per chromosome (default: NULL)
- position
"bp" or "Mb" (default: "bp")
- r2_threshold
r² value for decay distance calculation (default: 0.2),
- xlim_range
Zoom in range on the x axis, default is c(0, 20). Set to NULL to see entire plot.
- show_vline
show a vertical line intersecting the x-axis at the the half-decay distance (default, TRUE)
- show_hline
show a horizontal line intersecting the y-axis at the selected r² threshold (default, TRUE)
- seed
Optional integer seed to make the random sampling reproducible. Default 123.
Value
A list containing:
- plot
ggplot2 object with points and fitted curve (or NULL if fit fails)
- half_decay_dist
Distance where r² reaches the specified threshold (or NA if not estimable)
- model
Fitted nls object (or NULL if fitting failed)
Details
An asymptotic regression model is fit using SSasymp. The distance where r² reaches a specified threshold (e.g., 0.1 or 0.2) is calculated and returned. This function calculates r² by chromosome and follows the general logic of MapRtools::LD.plot but uses the function SSasymp to model the non-linear LD decay. This function will give you a quick and simple estimate of LD. For more sophisticated LD functions see David Gerard ldfast()
It processes each chromosome separately to compute all pairwise r² values within chromosomes, then it combines the results across all chromosomes into a single dataset, and fits one global LD decay model to the pooled data. No cross-chromosome marker pairs are considered.
If the user is interested in LD per chromosome, provide a map and data file with only data pertaining to the chromosome of interest.