Get smoothed f2-statistics — qpfstats • admixtools

This function returns an array of (pseudo-) f2-statistics which are computed by taking into account other f2-, f3-, and f4-statistics. The advantage of doing that is that f3- and f4-statistics computed from these smoothed f2-statistics can be more accurate for populations with large amounts of missing data. The function uses SNPs which are missing in some populations in a manner which tends to introduce less bias than setting maxmiss to values greater than 0.

qpfstats(
  pref,
  pops,
  include_f2 = TRUE,
  include_f3 = TRUE,
  include_f4 = TRUE,
  verbose = TRUE
)

Arguments

pref: Prefix of genotype files
pops: Populations for which to compute f2-statistics
include_f2: Should f2-statistics be used to get smoothed f2-statistics? If include_f2 is a positive integer, it specifies how many randomly chosen f2-statistics should be used.
include_f3: Should f3-statistics be used to get smoothed f2-statistics? If include_f3 is a positive integer, it specifies how many randomly chosen f3-statistics should be used.
include_f4: Should f4-statistics be used to get smoothed f2-statistics? If include_f4 is a positive integer, it specifies how many randomly chosen f4-statistics should be used.

Value

A 3d-array of smoothed f2-statistics

Examples

if (FALSE) {
f2_blocks = qpfstats(geno_prefix, mypops)
}