Additional file 1: of Detailed prediction of protein sub-nuclear localization

Figure S1. Effect of E-value thresholds on combined prediction of traveler proteins. Figure S2. Composition of sub-nuclear compartments in the human, chimp, mouse and yeast proteome and LocNuclei’s development set. Table S1. Composition of the sub-nuclear development set for LocNuclei. Table S2. LocNuclei confusion matrix for homology-based inference and machine learning prediction. Table S3. Comparison between LocNuclei and LocNuclei-NSort. Table S4. Euclidean distance between organisms based on subnuclear location spectra predicted with SVM Profile Kernel (de novo) or homology-based inference. Table S5. Top ten statistically enriched GO terms for each sub-nuclear compartment. Table S6. Normalization of sub-nuclear localization terms. Table S7. Chosen hyperparameters for the 14 different SVM Profile Kernels. (DOCX 43170 kb