Bioinformatics Analysis of Humoral Immune Response and Protein Microarray for Biomarker Discovery.

Immune Response;Protein Microarray;Bioinformatics;Science (General);Health Sciences;Science;Bioinformatics

Vuong, Huy Q.Pennathur, Subramaniam ;

University of Michigan

关键词: Immune Response;    Protein Microarray;    Bioinformatics;    Science (General);    Health Sciences;    Science;    Bioinformatics;   

Others  :  https://deepblue.lib.umich.edu/bitstream/handle/2027.42/89673/huyvuong_1.pdf?sequence=1&isAllowed=y

瑞士|英语

来源: The Illinois Digital Environment for Access to Learning and Scholarship

【 摘 要 】

Early detection is the best defense which could significantly improve the cancer survival rate in several cancers including melanoma and pancreatic cancer. A promising approach to discover biomarkers for early detection involves using the humoral immune response against tumor proteins. Together with advances in proteomics, in particular a high throughput protein microarray platform, humoral immune response studies have enabled a breakthrough in developing global screening of the highly complex plasma proteome for biomarkers for early detection. In this dissertation, we attempt to integrate proteomics and bioinformatics approaches to analyze signals from protein microarray data for the reliable identification of differentially expressed proteins under different biological conditions. First, we present a study comparing outlier-based to traditional mean-based approaches in differential expression analysis with applications in protein microarray data in heterogeneous diseases. Our investigation uses a glycoproteomics dataset from a melanoma study, an original simulation-based approach to benchmarking, and a new data visualization technique to assess the potential for methods that explicitly target heterogeneous patterns of differential expression to give improved performance relative to traditional approaches based on group-wise comparison of means. Results include identifications of 1 significant feature using outlier statistics and 15 significant features using t-statistics from a melanoma dataset of 43 samples and 47 features. Next, we apply the outlier strategy to a protein microarray dataset from a pancreatic cancer study with sera from 37 pancreatic cancers, 24 chronic pancreatitis and 23 normal to identify protein biomarkers that are differentially expressed in only a subset of cancer samples. Three protein markers exhibiting outlier patterns exclusive to cancer sera and no outliers in normal sera are identified by mass spectrometry and confirmed by a follow-up study with an independent dataset. The next study presents the application of a label-free quantitation approach for measuring changes in protein abundance level associated with phosphorylation, a major mechanism of tumorigenesis. Results include identifications of differentially expressed post-translational modified proteins such as phosphorylated lamin-A/C, isoforms-A and GTPase-activating protein binding protein-1 in pancreatic cancer. Together, this dissertation contributes two approaches to biomarker discovery using protein microarrays and the humoral immune response and LC-MS/MS and label-free quantitation.