Performance and Scalability of Discriminative Metrics for Comparative Gene Identification in 12 Drosophila Genomes
(A) Multidimensional scaling (MDS) visualization in which each point represents a metric and the distance between any two points approximately represents their dissimilarity, measured as 1-(rank correlation of the scores of the known exons). Hybrid metrics appear closer to the center, suggesting that they successfully combine distinct information from the individual metrics. (B) ROC curves showing the performance of two hybrid metrics created by combining five comparative and single-sequence metrics using Linear Discriminant Analysis (LDA) or a Support Vector Machine (SVM). The hybrid metrics outperformed all of their input metrics.