Enrichment Analysis In Predicting The Upstream Signal Transduction Pathway Through Expression Study

The study in signal transduction pathway was a centurial interest as a main concept in researching the mechanism of biology process. It was also acknowledged as a leading signal communication path among the multiple cells. Complex biological process such as proliferation, migration and apoptosis are generally regulated through responses of cells to stimuli in their environment. Meanwhile, plentiful of experiment results have confirmed a big role in the disease that the aberration of the pathway caused the occurrence of multi-type cancers. One of the widely profits in drug design is inhibited the cancer target genes through developing the signal pathway suppression, which was effective in clinical trials.Interestingly, the complex characteristic of signal pathway makes the literature work difficultly. At first, multiple molecular components involved in a path route served various relationships which can category in physical interactions, enzyme catalysis and transport. Even if in the same signal pathway, the biology switches play the different function under some certain conditions leading to specific biology function. Meanwhile, each pathway was necessarily cooperated with another signal pathway worked as a module to influence the biology process. That is to say, not only the construction of the signal pathway but also the module worked in a process, it is too complex to dig the regulation principle.In this thesis, we proposed a mature strategy, aiming at understanding the upstream signal pathway simulation at the microarray level in human. Firstly, based on the knowledge from integrating the database information, defined a list of "signature genes" reporting the reaction of biological experiments. Then calculate the weight scores with the multilevel information network following the similar network-based topology analysis model. Also, the cross-talk among the signal pathways will be discussed in real lung cancer data application. To better understanding this problem, the approach was developed in plenty text-mining, systematic network study and weight functional focused.