Multi-Scale Computational Analyses of Transcriptional Regulation

Table 1. Identified Pol II pausing sites in termination and promoter-proximal regions. Table 2. Architecture of the deep learning model implemented in Chapter 1. Table 3. Sum importance scores aligned to Pol II pausing sites identified through pentamer disruption analyses using the terminator and promoter models. Table 4. Genomic features surrounding terminator pausing sites used in subtype embedding analyses. Table 5. Annotation of the common zones for IRT and control genes, and the extended zones specific to IRT genes. Table 6. The stistatics of developmental lineages profiled by MetMap. Table 7. The differential gene expression analyseds performed in this study. Table 8. The coexpressed subGO modules identified in this study and the correlation between the subGO module expression and metastatic potentials across lineage types in Figure 10A. Table 9. The performance of random forest regression models across different metastatic routes. The Pearson correlation coefficient values between the predicted and observed values, as well as P-values are shown. Table 10. Jaccard index, Sørensen-Dice coefficient and max overlapping results between 230 subGOs and GO annotated child terms. Table 11. Node and edge features of network 1-7 from STRING analyses. Table 12. The drug response indicated by the Pearson correlation coefficient values and regression P-values between the IC50 values and the network expression.