As dependence has been an important theme in statistics for many years, there is large amount of literature in genetics as well as in statistics to tackle the problem. For example, the maximum likelihood or Bayes estimation of allele frequencies in relatives (Boehnke 1991; Thomas and Camp, 2006; Coram and Tang, 2007); the use of principal components or eigenvectors to identify clusters of samples (Price et al., 2006; Patterson et al., 2006), or the reduction of effective number of markers in a linkage disequilibrium block (Cheverud, 2001; Nyholt, 2004); sample weighting to suppress contributions from correlated samples (Broman, 2001; Browning et al., 2005), etc.
