System Biology Approach for Signaling Transduction Study of Complex Phenotypes

点击翻译按钮获取中文摘要

DESCRIPTION (provided by applicant): The primary goal of the proposed study is to develop a software package, Signal Transduction Network Analyzer (STkAnalyzer), to identify signal pathways and pathway signatures that are related to diseases with complex phenotype. We will use the four myelodysplastic syndromes (MDS) phenotypes as the prototype of disease to test the performance of this package by integrating high-throughput genome-wide profiling using single-nucleotide polymorphism (SNP) array, gene expression arrays, microRNA array, and publically available Kyoto Encyclopedia of Genes and Genomes (KEGG) and protein-protein interaction (PPI) databases. The frequency and incidence of MDSs is increasing in the U.S. population but the diagnosis of MDSs patients has not shown any significant improvement over the last decade. One major cause of the latter phenomenon is the lack of methodologies to accurately finding the pathways and biomarkers for MDSs at an early stage. Dr. Chang's group in The Methodist Hospital is studying a cohort of more than 300 well-characterized MDS patients. The MDS is characterized by very complex phenotypes with main categories include refractory anemia (RA), RA with ringed sideroblasts (RARS), refractory cytopenia with multi-lineage dysplasia (RCMD), RA with excess blasts (RAEB). Although MDS was used as the prototype of disease for this proposal, the package developed will be applicable to multiple diseases with complex phenotypes such as cancers, diabetes and so on. The impact of the package is tremendous in system biology. Detecting chromosomal abnormality can identify the candidate genetic alterations which may cause the transformation of the hematopoietic stem cells. But it cannot answer which of these candidate genes are the true causal genes of MDS phenotypes and how these genes cause MDS phenotypes. Similarly, comparison of gene expression profiles between disease samples and normal samples could identify which genes are active in disease tissue and which genes are inactive. However, it cannot discriminate which genes are the causes and which genes are the results. These questions are extremely important and the answers will shape our basic view of the molecular mechanism of MDS phenotypes and influences how to design and develop new strategies for diagnosis, treat and prevent MDS. The recent availability of large protein-protein interaction, protein-DNA interaction data, and the expression quantitative trait loci (eQTL) mapping techniques provides a means to address these issues. Hence we propose to identify signal pathways that are perturbed by susceptibility loci and that in turn lead to the four MDS phenotypes. The major technological contributions in this package (STkAnalyzer) are in four: first, a novel Conditional Random Pattern approach is developed for amplified SNParray copy number estimation and LOH detection; second, eQTL mapping is proposed to associate the genotyping data and mRNA; third a significance analysis of microRNA-mRNA targeting (SAMiMT) is proposed to integrate mRNA and microRNA arrays, and finally a Diffusion Mapping or Semi-Group approach is proposed for inferring signal transduction network and biomarker motif (biomarker pattern or pathway signature) to unravel the underlying mechanism how the eQTLs lead to the MDS pathogenesis.