Holmquist R, Jukes T H
J Mol Evol. 1981;18(1):47-59. doi: 10.1007/BF01733211.
The recent evaluation by Fitch (1980) of REH theory for macromolecular divergence is a severely erroneous and distorted analysis of our work over the past decade. We reply to those distortions here. At present, there is no factual basis for believing Fitch's assessment that corrections which move evolutionary estimates of total mutations fixed closer to the true distance must do so at the expense of an increased variance sufficient to compromise the value of the improvement. By direct calculation the variance in the estimates of total mutations fixed given by REH theory is comparable to that of other models now in the literature for the case in which genetic events are equiprobable. A general argument is given that suggests that, as we consider more and more carefully the selective, functional, and structural constraints on the evolution of genes and proteins, this variance may be expected to decrease toward a lower bound.
惠誉（1980年）最近对大分子分歧的相对熵假说（REH）理论的评估，是对我们过去十年工作的严重错误和扭曲的分析。我们在此回应这些歪曲。目前，没有事实依据支持惠誉的评估，即那些使固定总突变的进化估计值更接近真实距离的校正，必然是以增加方差为代价，而这种方差的增加足以损害改进的价值。通过直接计算，在遗传事件等概率的情况下，REH理论给出的固定总突变估计值的方差与文献中其他模型的方差相当。一个一般性的论点表明，随着我们越来越仔细地考虑基因和蛋白质进化中的选择、功能和结构限制，这种方差可能会朝着一个下限减小。
