1.A Study On Expression Signature Prognostic For Survival Of Lung Adenocarcinoma From MRNA Profiling In Human Lung Development 2.Power Of Deep Sequencing And Agilent Microarray For Gene Expression Profiling Study

This PhD degree project is composed of two parts of study.Part I A Study on Expression Signature Prognostic for Survival of Lung Adenocarcinoma from mRNA Profiling in Human Lung DevelopmentPurpose:It is demonstrated that embryonic development to some extent resembles tumorigenesis. Thus, it would be helpful to better understand molecular feature of tumors which is characterized by gene expression signatures in embryonic developmental process. In this study, we tried to identify the development signature as being survival related gene expression profile by following observations:1) the mRNA expression profile in human embryonic development; 2) similarity comparison of gene expression signature between human embryonic development and primary non-small cell lung cancer (NSCLC); and 3) the association between embryonic development gene signature and risk stratification of NSCLC. Results:Two sets of genes which exhibited "X" expression pattern were identified. Gene ontology analysis showed that the genes activated in early stage of development associated with cell proliferation, while another set of genes up-regulated in late stage was implicated in cell-cell, cell-matrix communication and cell membrane maturation. By gene expression profiling, we defined a human fetal lung developmental landscape on which the gene expression data from NSCLC were projected. It was found that the gene signatures from adenocarcinoma and squamous cell carcinoma were similar to gene expression profiles in developing lung tissues. The expression of SMC4 represented gene set activated in early stage and gradually abated in mid-late stage of development was up-regulated in NSCLC tissue. Further, independent of TNM staging parameters, the expression of SMC4-gene set correlated with overall survival of lung adenocarcinoma patients, and the phenomenon was subsequently validated by multiple independent sample sets. However, there was no association between developmental signature and the prognosis of patients with lung squamous cell cancer. Conclusion:Part of lung adenocarcinoma transcriptome is composed of developmental gene signatures and survival related malignant features may be reminiscent of embryonic behavior. Comprehensive analysis on development signatures of cancers would be a meaningful approach to uncover critical clues to tumorigenesis.Part II Power of Deep Sequencing and Agilent Microarray for Gene Expression Profiling StudyNext-generation sequencing based Digital Gene Expression tag profiling (DGE) has been used to study changes in gene expression profiling. To compare the quality of the data generated by microarray and DGE, we examined the gene expression profiles of an in vitro cell model with both platforms. In this study,17,362 and 15,938 genes were detected by microarray and DGE, respectively, with 13,221 overlapping genes. The correlation coefficients between the technical replicates were>0.99 and the detection variance was<9% for both platforms. The dynamic range of microarray was fixed with 4 orders of magnitude, while the dynamic range of DGE was extendable. The consistency of the two platforms was high, especially for those abundant genes. It was more difficult for the microarray to distinguish the expression variation of less abundant genes. Although microarrays might be eventually replaced by DGE or transcriptome sequencing (RNA-seq) in the near future, microarrays are still stable, practical and feasible, which renders the same useful for most biological researchers.