National Center for Biotechnology Information, National Library of Medicine, National Institutes of Health, Bethesda, Maryland, USA.
Laboratory of Cellular and Developmental Biology, National Institute of Diabetes and Digestive and Kidney Diseases, Bethesda, Maryland, USA.
J Comput Biol. 2023 Jan;30(1):21-40. doi: 10.1089/cmb.2022.0121. Epub 2022 Aug 26.
Gene expression evolution is typically modeled with the stochastic Ornstein-Uhlenbeck (OU) process. It has been suggested that the estimation of within-species variations using replicated data can increase the predictive power of such models, but this hypothesis has not been fully tested. We developed EvoGeneX, a computationally efficient implementation of the OU-based method that models within-species variation. Using extensive simulations, we show that modeling within-species variations and appropriate selection of species improve the performance of the model. Further, to facilitate a comparative analysis of expression evolution, we introduce a formal measure of evolutionary expression divergence for a group of genes using the rate and the asymptotic level of divergence. With these tools in hand, we performed the first-ever analysis of the evolution of gene expression across different body-parts, species, and sexes of the genus. We observed that genes with adaptive expression evolution tend to be body-part specific, whereas the genes with constrained evolution tend to be shared across body-parts. Among the neutrally evolving gene expression patterns, gonads in both sexes have higher expression divergence relative to other tissues and the male gonads have even higher divergence than the female gonads. Among the evolutionarily constrained genes, the gonads show different divergence patterns, where the male gonads, and not the female gonads, show less constrained divergence than other body-parts. Finally, we show interesting examples of adaptive expression evolution, including adaptation of odor binding proteins.
基因表达进化通常采用随机的奥恩斯坦-乌伦贝克（Ornstein-Uhlenbeck，OU）过程进行建模。有人认为，使用重复数据来估计种内变异可以提高此类模型的预测能力，但这一假设尚未得到充分验证。我们开发了 EvoGeneX，这是一种基于 OU 的方法的计算效率实现，可对种内变异进行建模。通过广泛的模拟，我们表明，对种内变异进行建模和对物种进行适当选择可以提高模型的性能。此外，为了便于对表达进化进行比较分析，我们使用速率和渐近水平的分歧，为一组基因引入了一种正式的进化表达分歧度量。有了这些工具，我们首次对 属不同身体部位、物种和性别的基因表达进化进行了分析。我们观察到，具有适应性表达进化的基因往往具有特定的身体部位特异性，而具有约束进化的基因则倾向于在身体部位之间共享。在中性进化的基因表达模式中，两性的性腺相对于其他组织具有更高的表达分歧，而雄性性腺的分歧甚至高于雌性性腺。在进化上受约束的基因中，性腺表现出不同的分歧模式，其中雄性性腺而不是雌性性腺，其分歧比其他身体部位受约束更小。最后，我们展示了一些有趣的适应性表达进化的例子，包括气味结合蛋白的适应。
