Estimated cM from kinship coefficient

[stephenturner]

See footnote on page 5:

• Expected cM = min(3560, 4 × max(0, kinship_coeff) × 3560

snp-typing-uas-kinship-estimation-gedmatch-pro-tech-note-vd2020058-a.pdf

library(skater)
library(dplyr)
dibble(9) %>%
  mutate(cm=pmin(3560, 4*pmax(0, k)*3560))
#> # A tibble: 11 x 5
#>    degree        k         l        u     cm
#>     <int>    <dbl>     <dbl>    <dbl>  <dbl>
#>  1      0 0.5       0.354    1        3560  
#>  2      1 0.25      0.177    0.354    3560  
#>  3      2 0.125     0.0884   0.177    1780  
#>  4      3 0.0625    0.0442   0.0884    890  
#>  5      4 0.0312    0.0221   0.0442    445  
#>  6      5 0.0156    0.0110   0.0221    222. 
#>  7      6 0.00781   0.00552  0.0110    111. 
#>  8      7 0.00391   0.00276  0.00552    55.6
#>  9      8 0.00195   0.00138  0.00276    27.8
#> 10      9 0.000977  0.000691 0.00138    13.9
#> 11     NA 0        -1        0.000691    0

Estimated cM appear to line up with expected relationship types from shared cM project and AncestryDNA matching white paper.