Discriminating and locating character covariance: An application of permutation tail probability (PTP) analyses

The ability of permutation tail probability (PTP) analyses to discriminate between character covariance and noise is investigated with both hypothetic al and published data sets. PTP is shown to be a powerful tool, not only fo r detecting character covariance, but also for locating that covariance on trees. PTP is especially useful for evaluating DNA sequence data that may h ave a high level of homoplasy. A three-step PTP procedure for locating cova ried characters is presented.