Chronograms—phylogenies with branch lengths proportional to time—represent key data on timing of evolutionary events, allowing us to study natural processes in many areas of biological research. Chronograms also provide valuable information that can be used for education, science communication, and conservation policy decisions. Yet, achieving a high-quality reconstruction of a chronogram is a difficult and resource-consuming task. Here we present DateLife, a phylogenetic software implemented as an R package and an R Shiny web application available at www.datelife.org, that provides services for efficient and easy discovery, summary, reuse, and reanalysis of node age data mined from a curated database of expert, peer-reviewed, and openly available chronograms. The main DateLife workflow starts with one or more scientific taxon names provided by a user. Names are processed and standardized to a unified taxonomy, allowing DateLife to run a name match across its local chronogram database that is curated from Open Tree of Life’s phylogenetic repository, and extract all chronograms that contain at least two queried taxon names, along with their metadata. Finally, node ages from matching chronograms are mapped using the congruification algorithm to corresponding nodes on a tree topology, either extracted from Open Tree of Life’s synthetic phylogeny or one provided by the user. Congruified node ages are used as secondary calibrations to date the chosen topology, with or without initial branch lengths, using different phylogenetic dating methods such as BLADJ, treePL, PATHd8, and MrBayes. We performed a cross-validation test to compare node ages resulting from a DateLife analysis (i.e, phylogenetic dating using secondary calibrations) to those from the original chronograms (i.e, obtained with primary calibrations), and found that DateLife’s node age estimates are consistent with the age estimates from the original chronograms, with the largest variation in ages occurring around topologically deeper nodes. Because the results from any software for scientific analysis can only be as good as the data used as input, we highlight the importance of considering the results of a DateLife analysis in the context of the input chronograms. DateLife can help to increase awareness of the existing disparities among alternative hypotheses of dates for the same diversification events, and to support exploration of the effect of alternative chronogram hypotheses on downstream analyses, providing a framework for a more informed interpretation of evolutionary results.

中文翻译：

---

DateLife：利用数据库和分析工具揭示过时的生命之树


计时图——分支长度与时间成正比的系统发育——代表了进化事件时间的关键数据，使我们能够研究生物学研究许多领域的自然过程。Chronograms 还提供了可用于教育、科学传播和保护政策决策的宝贵信息。然而，实现天文台表的高质量重建是一项艰巨且耗费资源的任务。在这里，我们介绍了 DateLife，这是一个作为 R 包实现的系统发育软件，www.datelife.org 上提供了一个 R Shiny Web 应用程序，它为高效、轻松地发现、总结、重用和重新分析从专家、同行评审和公开可用的计时图的精选数据库中挖掘的节点年龄数据提供服务。主 DateLife 工作流从用户提供的一个或多个科学分类名称开始。名称经过处理并标准化为统一的分类法，允许 DateLife 在其本地计时图数据库中运行名称匹配，该数据库是从 Open Tree of Life 的系统发育存储库中策划的，并提取包含至少两个查询的分类名称的所有计时图及其元数据。最后，使用同构算法将匹配计时图中的节点年龄映射到树形拓扑上的相应节点，这些节点可以从 Open Tree of Life 的合成系统发育中提取，也可以由用户提供。一致的节点年龄用作二次校准，以使用不同的系统发育测年方法（如 BLADJ、treePL、PATHd8 和 MrBayes）对所选拓扑进行测年，无论是否具有初始分支长度。我们进行了交叉验证测试，将 DateLife 分析（即使用二次校准的系统发育测年）得出的节点年龄与原始计时图 （i.e，通过初级校准获得），并发现 DateLife 的节点年龄估计值与原始计时图的年龄估计值一致，年龄的最大变化发生在拓扑更深的节点周围。由于任何用于科学分析的软件的结果都只能与用作输入的数据一样好，因此我们强调在输入计时器的背景下考虑 DateLife 分析结果的重要性。DateLife 可以帮助提高对相同多元化事件的日期替代假设之间现有差异的认识，并支持探索替代时序图假设对下游分析的影响，为更明智地解释进化结果提供框架。