The global distribution of known and undiscovered ant biodiversity

This is the Data S1 supplemental archive for the paper "The global distribution of known and undiscovered ant biodiversity" in Science Advances by Kass et al. All methodological details can be found in the Materials and Methods section of the paper. The archive contains three data directory archives, programming code archived as software with Zenodo, supplementary figures and model metadata archived as supplemental information with Zenodo, and a master README file. Please see more detailed descriptions below. Data main_analysis_data.zip: The directory "main_analysis_data", archived in the "main_analysis_data.zip" file, contains all the core data used in the analyses described in the paper, except those data described below that were too big to include in one compressed file. This includes the Global Ant Biodiversity Informatics (GABI) data, both raw and after cleaning and geocoding, all diversity estimate raster data shown in the figures of the paper, and other related data. Please consult "main_analysis_data/README_main_analysis_data.txt" for more details. NOTE: If researchers encounter issues reproducing results with the code provided, please contact <[e-mail]>.results_sp_gen_data_add.zip: Additional results for species and genera, including individual range estimates from polygons and species distribution models, individual datasets used for modeling, and intermediate occurrence data subsets for the geocoding analysis, are found in "results_species_add" and "results_genus_add", respectively. Intermediate occurrence data subsets for the geocoding analysis are also archived in "processing_data_add".random_forest_add.zip: The fitted Random Forest models used to make predictions of unknown diversity centers under a global high-sampling scenario, with variable importance calculated.Please see more details in README_MASTER.txt. SoftwareKass_et_al_2022_SciAdv_prog_code.zip: A simple package that contains all the R and Python scripts used to conduct the analysis and generate the figures in the paper. This folder has its own separate README with more details. The easiest way to run code is to open the .Rproj file in RStudio and press the "Install and Restart" button under the "Build" tab in the Environment frame -- this installs and loads the package and thus makes all functions available in the programming environment. The main analysis script is located in analysis/main_analysis.R.Supplemental Informationout_irs60K.zip: Example figure plots made by Kass_et_al_2022_SciAdv_prog_code/analysis/figures.R. The figures displayed in the paper were made with ArcGIS using the same underlying data. Fig. 3 is not represented here because it was made with data processed in ArcGIS, but these processed files can be found in main_analysis_data/overlays/for_fig3.ODMAP_model_metadata.csv: Metadata for Maxent species distribution models and Random Forest models structured according to the ODMAP (Overview, Data, Model, Assessment and Prediction) framework formalized by Zurell et al. (2020) [https://doi.org/10.1111/ecog.04960]. This metadata was created using the shiny app located at https://odmap.wsl.ch/ and was edited lightly by hand to include some extra detail.

本档案为Kass等人发表于《科学·进展》（Science Advances）的论文《已知与未发现蚂蚁生物多样性的全球分布》配套的Data S1补充档案。所有方法学细节均可在论文的“材料与方法”章节中查阅。本档案包含三份数据目录压缩包、存储于Zenodo的编程代码软件存档、作为补充信息存储于Zenodo的补充图表与模型元数据，以及一份主README文件。下文将提供更详细的说明。

main_analysis_data.zip：该压缩包内的“main_analysis_data”目录包含了论文所述分析中使用的全部核心数据，除因体积过大无法纳入单个压缩文件的部分数据外。其中涵盖全球蚂蚁生物多样性信息学（Global Ant Biodiversity Informatics, GABI）的原始数据、经清洗与地理编码后的数据集、论文各图中展示的所有多样性估算栅格数据，以及其他相关数据。更多细节请参阅“main_analysis_data/README_main_analysis_data.txt”。

注意：若研究人员在使用提供的代码复现研究结果时遇到问题，请联系<[e-mail]>。

results_sp_gen_data_add.zip：物种与属级别的额外结果分别存储于“results_species_add”与“results_genus_add”目录中，包括基于多边形的单物种分布范围估算结果、物种分布模型、建模所用的单个数据集，以及地理编码分析所用的中间物种出现数据子集。地理编码分析所用的中间物种出现数据子集同时存档于“processing_data_add”目录。

random_forest_add.zip：包含已拟合的随机森林模型，用于在全球高采样情景下预测未知生物多样性热点区域，并计算了变量重要性。更多细节请参阅README_MASTER.txt。

软件包Kass_et_al_2022_SciAdv_prog_code.zip：该简易软件包包含了用于开展论文分析与生成图表的全部R与Python脚本。该文件夹自带独立的README文件以提供更多细节。运行代码的最简方式为在RStudio中打开.Rproj文件，并在环境面板的“构建”标签下点击“安装并重启”按钮——此操作将安装并加载该软件包，从而使编程环境中可调用所有函数。主分析脚本位于analysis/main_analysis.R。

补充信息：out_irs60K.zip：由Kass_et_al_2022_SciAdv_prog_code/analysis/figures.R生成的示例图表。论文中的图表是使用ArcGIS结合相同的基础数据制作而成。由于图3采用ArcGIS中处理后的数据生成，因此未在此处展示，但这些处理后的数据可于main_analysis_data/overlays/for_fig3中找到。

ODMAP_model_metadata.csv：按照Zurell等人2020年提出的ODMAP（Overview, Data, Model, Assessment and Prediction，概述、数据、模型、评估与预测）框架构建的Maxent物种分布模型与随机森林模型元数据[https://doi.org/10.1111/ecog.04960]。该元数据通过位于https://odmap.wsl.ch/的shiny应用生成，并经手动小幅编辑以补充部分额外细节。