Data for Error-correcting output codes and multi-view learning in the tissue of origin classification

As various modalities of genomic data are accumulating, methods to integrate across multi-omics datasets are becoming important. Error-correcting output codes (ECOC) is an ensemble learning strategy for solving a multiclass problem thru a decoding process that aggregates the predictions of multiple classifiers. Thus, it lends itself naturally to aggregating predictions across multiple views as well. We applied the ECOC to multi-view learning to see if this strategy can enhance classifier performance as compared to traditional techniques. We designed experiments to predict tissue types for hundreds of samples using measures of the transcriptome and methylome. We tested our ECOC design for multi-view learning, where the feature sets for RNA-Seq and DNA methylation were encoded separately and decoded together, to see if we could achieve better performance as compared to the traditional uses of feature sets. Our analyses revealed that multi-view ensemble ECOC achieved higher classifier performance in certain experimental designs. The novel multi-view ensemble ECOC method merits consideration by other researchers to potentially attain superior classification results.

随着基因组数据的多种模态不断累积，跨多组学数据集的整合方法正变得日益重要。错误纠正输出码（ECOC）是一种集成学习策略，通过解码过程聚合多个分类器的预测来解决多分类问题。因此，它天然适用于聚合多个视角的预测。我们应用ECOC于多视角学习，以探究该策略是否能够提升分类器的性能，相较于传统技术。我们设计了实验，利用转录组和甲基化组测量指标预测数百个样本的组织类型。我们对多视角学习的ECOC设计进行了测试，其中RNA-Seq和DNA甲基化的特征集分别编码并共同解码，以观察我们是否能够实现相较于传统特征集使用的更好性能。我们的分析揭示了，在特定实验设计中，多视角集成ECOC实现了更高的分类器性能。新颖的多视角集成ECOC方法值得其他研究人员考虑，以可能获得更优的分类结果。