STS Postdoctoral Fellowship: The impact of mathematics on inferring classifications and phylogenies

本提案将探讨计算生物学中数学建模的生物学基础。系统生物学家对生物体进行分类和系统发育。这样做是基于这样一种想法，即添加更多的数据将使他们的结果在统计上趋于一致。然而，这种趋同的生物学基础差异很大，并且在推理方法和生物体之间存在很大争议。该项目将采用一种混合方法，将哲学分析与过去六十年来系统学中统计方法接受的历史研究相结合，包括数字分类学、分类学和最大似然方法，来解决系统生物学中统计趋同的问题。这项研究将强调统计方法的使用和价值如何在系统分类学的各个领域，如微生物学、动物学和古生物学中有所不同。研究人员将开发计算方法来追踪已发表的期刊文章的接收情况，以便进行全面的比较分析。该项目将使用开创性的OCHRE数据库系统，该系统在数字人文学科中提供了一种创新的数据结构，可以支持随时间变化的分析类别（半结构化数据库而不是关系数据库）。本项目是一个跨学科的合作项目，将生物学、科学哲学和科学史结合起来，研究系统学中统计方法的优势和局限性。菲尔德博物馆是这次合作的理想地点，因为它在系统学方面的研究卓越而多样，包括生物多样性研究整合数据的前沿项目，生命之树项目的分子技术，以及系统学的历史和哲学。这项研究的重点是统计学对系统分类学研究的约束，它将打开一个新的认识论视角，把个性作为一个互补的问题，特别是在分类或系统发育中的结构要素个体化以及有机体的个体化特征方面。研究不同的系统学领域是如何接受和利用统计方法的，将极大地扩展过去六十年来微生物学和植物学理论进展的历史知识。此外，该项目将开发研究工具，用于研究生物学和其他科学的数学化，这是科学史和科学哲学中日益重要的问题。例如，提出的研究结果之一将是计算方法，用于分析数学化对系统学研究问题之间的结构关系的变革性影响。潜在的更广泛的影响利用综合的跨学科方法来解决当今生物学中的一个关键实际问题，这项研究的结果将跨越边界，影响生物学和数字人文学科其他领域的科学政策。该项目将对统计趋同的基础进行有针对性的评估，这与目前决策者和生物学家之间关于数学和高通量数据收集技术价值的辩论有关。研究结果将在跨学科期刊和会议上发表，这些期刊和会议的听众中经常包括生物学家。在该项目中开发的新计算方法将作为Java模块并入OCHRE系统，该系统可用于学术研究，并具有基于网络的图形用户界面。新的计算方法和相关结果也将在数字人文和数字历史与科学哲学研讨会上发表。

IntroductionThis proposal will examine the biological basis of mathematical modeling in computational biology. Systematic biologists make classifications and phylogenies of organisms. Doing so depends on the idea that adding more data will cause their results to converge statistically. However, the biological basis for this convergence varies considerably and is heavily contested across inference methods and organisms. This project will approach the problem of justifying statistical convergence in systematic biology by adopting a hybrid approach that integrates philosophical analysis with a historical study of the reception of statistical methods in systematics over the past sixty years, including numerical taxonomy, cladistics, and maximum likelihood methods. The study will emphasize how the use and value of statistical methods varied across areas of systematics, such as microbiology, zoology, and paleontology. The researchers will develop computational methods for tracing reception in published journal articles in order to achieve a comprehensive, comparative analysis. The project will use the path-breaking OCHRE database system, which provides an innovative data structure in the digital humanities that can support changing analytic categories over time (a semi-structured instead of relational database). Intellectual MeritThis project is an interdisciplinary collaboration integrating biology, philosophy of science, and history of science to study the strengths and limitations of statistical methods in systematics. The Field Museum is an ideal location for this collaboration because of the excellence and diversity of its research in systematics, including cutting-edge projects on integrating data for biodiversity studies, on molecular techniques for the Tree of Life project, and on the history and philosophy of systematics. The study, with its focus on statistics as a constraint on research in systematics, will open a new epistemological perspective on individuality as a complementary problem, especially with regards to individuating structural elements in classifications or phylogenies and with regards to individuating characters in organisms. Examining how different areas of systematics received and appropriated statistical methods will greatly expand current historical knowledge of theoretical advances in microbiology and botany over the past sixty years. In addition, the project will develop research tools for investigating the mathematization of biology and other sciences, an increasingly important problem in the history and philosophy of science. For example, one result of the proposed research will be computational methods for analyzing the transformative effect mathematization had on the structural relations among research problems in systematics. Potential Broader ImpactUsing an integrated interdisciplinary approach to address a key practical problem in biology today, the results of this research will cross boundaries to impact science policy in biology and other fields in the digital humanities. The project will produce a targeted evaluation of the basis of statistical convergence that is relevant to current debates among policy makers and biologists about the value of mathematics and high-throughput data gathering techniques. Results will be disseminated in interdisciplinary journals and conferences that regularly include biologists in their audience. New computational methods developed in the project will be incorporated as Java modules in the OCHRE system, which is made available for use in academic research and has a web-based graphical user interface. The new computational methods and associated results will also be presented at workshops on digital humanities and digital history and philosophy of science.