# popkin_A_min_subpops: Estimate the minimum expected value of a matrix 'A' using... In popkin: Estimate Kinship and FST under Arbitrary Population Structure

 popkin_A_min_subpops R Documentation

## Estimate the minimum expected value of a matrix `A` using subpopulations

### Description

This function averages the values of a square matrix `A` between every subpopulation pair and returns the minimum of these averages. If no subpopulation partition is provided, the function returns the minimum value of `A` excluding the diagonal, to agree when the code treats each individual as a subpopulation. The return value can be used to adjust an `A` matrix to yield the kinship matrix.

### Usage

```popkin_A_min_subpops(A, subpops = NULL)
```

### Arguments

 `A` A symmetric `n`-by-`n` matrix with values between every individual pair, including self comparisons. `subpops` A length-`n` vector of subpopulation assignments for each individual. If missing, every individual is treated as a different subpopulation.

### Value

The minimum of the average between-subpopulation `A` values, which estimates the minimum expected value of `A`

`popkin_A()` to generate the `A` matrix usually inputted into this function (`popkin_A_min_subpops`). `popkin()` is the wrapper function around both of these.

`avg_kinship_subpops()` for the full matrix of mean kinship values between subpopulations.

### Examples

```# Construct toy data
X <- matrix(c(0,1,2,1,0,1,1,0,2), nrow=3, byrow=TRUE) # genotype matrix
subpops <- c(1,1,2) # subpopulation assignments for individuals

# NOTE: for BED-formatted input, use BEDMatrix!
# "file" is path to BED file (excluding .bed extension)
## library(BEDMatrix)
## X <- BEDMatrix(file) # load genotype matrix object

# calculate A from genotypes
A <- popkin_A(X)\$A

# the recommended form using appropriate subpopulation labels
A_min_est <- popkin_A_min_subpops( A, subpops )

# this recovers the popkin estimate
kinship <- 1 - A / A_min_est
stopifnot( kinship == popkin( X, subpops ) )

# a simple default for exploratory analysis, equals min( A ) for correctly-calculated A
A_min_est <- popkin_A_min_subpops( A )
stopifnot( A_min_est == min( A ) )

```

popkin documentation built on Jan. 7, 2023, 1:26 a.m.