Counting chickens before they hatch: reciprocal consistency of calibration points for estimating divergence dates

There has been concern in the literature about the methodology of using secondary calibration timepoints when estimating evolutionary divergence dates. Such timepoints are divergence time estimates that have been derived from one molecular data set on the basis of a primary external calibration timepoint, and which are then used independently on a second data set. Logically, the primary and secondary calibration points must be mutually consistent, in the sense that it must be possible to predict each time point from the other. However, the attempt by Shaul and Graur (2002, Gene 300: 59–61) to assess the reliability of secondary timepoints is flawed because they presented time estimates without presenting confidence intervals on those estimates, and so it was not possible to make any explicit hypothesis tests of divergence times. Also, they inappropriately excluded some of the data, which leads to a very biased estimate of one of the divergence times. Here, I present a re-analysis of the same data set, with more appropriate methodology, and come to the conclusion that no inconsistencies are involved. However, it is clear from the analysis that molecular data often have such large confidence intervals that they are uninformative, and thus cannot be used for reliable hypothesis tests.