Can indirect tests detect a known recombination event in human mtDNA?

Whether human mitochondrial DNA (mtDNA) recombines sufficiently to influence its evolution, evolutionary analysis, and disease etiology, remains equivocal. Overall, evidence from indirect studies of population genetic data suggests that recombination is not occurring at detectable levels. This may be explained by no, or low, recombination or, alternatively, current indirect tests may be incapable of detecting recombination in human mtDNA. To investigate the latter, we have tested whether six well-established indirect tests of recombination could detect recombination in a human mtDNA data set, in which its occurrence had been empirically confirmed. Three showed statistical evidence for recombination (r(2) vs. distance, the Homoplasy test, Neighborhood Similarity Score), and three did not (D' vs. distance, Max Chi Squared, Pairwise Homoplasy Index). Possible reasons for detection failure are discussed. Further, evidence from earlier studies suggesting a lack of recombination in mtDNA in humans is reconsidered, taking into account the appropriateness of the tests used, based on our new findings.