Spearman rank correlations after soft integration of other data sets.

The expression data are obtained from the Rosetta Inpharmatics and ncbi omnibus, and integrated into the inference process by more terms in the objective function. The TF-binding data come from Yeastract and form priors for the penalty term, making it more probable that genes which are co-regulated should act as predictors for each other. Both data sets are only included to the extent the cross-validation procedure allows.

本研究的基因表达谱数据获取自Rosetta Inpharmatics与NCBI基因表达综合数据库（ncbi omnibus），并通过目标函数中的额外项整合至模型推断流程中。转录因子结合（Transcription Factor Binding, TF-binding）数据来源于Yeastract数据库，作为惩罚项的先验信息，使得共调控基因更有可能互为预测因子。两类数据集仅在交叉验证流程允许的范围内被纳入分析。