Performance of the different types of gene and miRNA models.

Number of significant models, scaled mean squared prediction error from cross-validation (cv-SMSE), 0.05 and 0.95 quantile of the differences in gene-wise correlation between original and predicted gene expression, and range and median of number of predictors for each model type.该数据集包含以下统计量：具有统计显著性的模型数量、基于交叉验证的标准化均方预测误差（cv-SMSE）、原始基因表达与预测基因表达的逐基因相关性差异的0.05分位数与0.95分位数，以及各模型类型的预测变量数量的极差与中位数。