Additional file 1 of Variables for habitat and vertebrate hosts of Ixodes scapularis are the best ecological predictors of the spatial spread of Lyme disease in the United States (2010–2019)

Additional file 1 (Additional file 1: Datafile S1. An Excel spreadsheet file including (a) a description of each explanatory variable, (b) the complete raw data used in the development of this study, (c) the population and case report of each county for the complete study period (obtained from the CDC and US Census Office). The columns of sheets b and c have a numeric indication of the year. Additional file 2: Datafile S2. A script including the development of the model in Python language, running under the umbrella of Orange data mining software (open-access software). To run this script, a basic knowledge of Python language is required. Additional file 3: Table S1. The confusion matrixes of the number of correctly or incorrectly allocated counties regarding the actual cases classes (rows) versus those predicted by the random forest and gradient boosting models (columns); results include the performance metric results for the individually modeled years 2010, 2013, 2016, and 2019. Values listed correspond to the proportion (%) of counties correctly identified by the models with the reported Lyme disease incidence class for each year. ∑ corresponds to the number of counties allocated for each incidence class. Additional file 4: Table S2. Tabular results for the associations between abiotic explanatory variables for landscape, vegetation, and climate with the 1322 US counties’ Lyme disease incidence classes in the complete 2010–2019 series.

附加文件1：数据文件S1：为Excel电子表格文件，包含(a)各解释变量（explanatory variable）的详细说明，(b)本研究开发过程中所使用的全部原始数据（raw data），(c)完整研究周期内各县级行政区的人口与病例报告数据（数据来源于美国疾病控制与预防中心（CDC）与美国人口普查办公室（US Census Office））。工作表b与c的列包含以数值形式标注的年份信息。
附加文件2：数据文件S2：为Python语言编写的模型开发脚本，运行于开源数据挖掘软件Orange（Orange data mining software）中。运行该脚本需具备基础的Python语言知识。
附加文件3：表S1：为混淆矩阵（confusion matrix），展示了基于实际病例类别（行）与随机森林（random forest）、梯度提升（gradient boosting）模型预测类别（列）的县级行政区分类正误情况；结果涵盖2010、2013、2016及2019各单独建模年份的性能指标（performance metric）结果。所列数值为各年份模型对莱姆病（Lyme disease）发病等级的县级行政区正确识别的占比（%）。符号∑代表各发病等级对应的县级行政区分配总数。
附加文件4：表S2：为表格结果，展示了2010-2019完整研究序列中，景观（landscape）、植被（vegetation）与气候（climate）相关的非生物解释变量（abiotic explanatory variable）与美国1322个县级行政区莱姆病（Lyme disease）发病等级之间的关联关系。