Transcriptional and splicing regulatory networks: Applications in cancer and stem cell differentiation

The complexity of the human regulatory networks makes it a great chanllenge in understanding them and delineating their implications for cancer and stem cell differentiation. In this study, automated systems that construct transcriptional and alternative splicing regulatory networks based on gene and exon microarray analysis, correlation method, bioinformatics tools and Bayesian integration were developed. Based on these regulatory networks, an automated nonlinear dynamical system analysis workflow was developed to delineate mechanisms of cancer onset and progression. Quantitative models to explain stem cell differentiation phenomenon based on the concepts of "symmetry-breaking instability" and "self-organization" were developed. A model delineating the mechanisms of Notch signaling in directing T/B cell differentiation and especially in T-helper (TH) (T H1/TH2) cell differentiation was developed. A group of genes support stem cell differentiation was revealed to be strongly over-represented in a database of cancer-related genes.