¹Center for the Study of Evolution, School of Biological Sciences, Biols, Sussex University, BN1 9QG Brighton, UK

Correspondence: G Piganeau, Center for the Study of Evolution, School of Biological Sciences, Biols, Sussex University, BN1 9QG Brighton, UK. E-mail: g.v.piganeau@sussex.ac.uk

Received 1 November 2002; Accepted 19 September 2003; Published online 28 January 2004.

Abstract

In an attempt to resolve the controversy about whether recombination occurs in human mtDNA, we have analysed three recently published data sets of complete mtDNA sequences along with 10 RFLP data sets. We have analysed the relationship between linkage disequilibrium (LD) and distance between sites under a variety of conditions using two measures of LD, r² and |D'|. We find that there is a negative correlation between r² and distance in the majority of data sets, but no overall trend for |D'|. Five out of six mtDNA sequence data sets show an excess of homoplasy, but this could be due to either recombination or hypervariable sites. Two additional recombination detection methods used, Geneconv and Maximum Chi-Square, showed nonsignificant results. The overall significance of these findings is hard to quantify because of nonindependence, but our results suggest a lack of evidence for recombination in human mtDNA.