Additional file 14 of Profiling non-coding RNA levels with clinical classifiers in pediatric Crohn’s disease

Additional file 14: Table S13. Using RandomForest, n = 274 ileal patients were randomly split into equal number of samples, n = 14 B1 and n = 14 B3. These were further aliquoted into test and train datasets for cross-validation of AUC with n = 13,777 ncRNAs, base mean greater than n > 10.00. The input included n = 14 B1 and n = 14 B3 used to test disease versus controls status by evaluating model and test for accuracy based on sensitivities and specificity.

补充文件14：表S13。本研究采用随机森林（RandomForest）算法，将274例回肠患者样本随机均分为两组，分别包含14份B1样本与14份B3样本。随后将上述样本进一步划分为训练集与测试集，以开展曲线下面积（AUC）的交叉验证分析，本次分析共纳入13777个非编码RNA（ncRNAs），其碱基均值大于10.00。本研究的输入数据集包含14份B1样本与14份B3样本，通过构建模型并基于灵敏度与特异度检验模型准确性，以验证疾病组与对照组的分类状态。