This subproject is one of many research subprojects utilizing the resources provided by a Center grant funded by NIH/NCRR. Primary support for the subproject and the subproject's principal investigator may have been provided by other sources, including other NIH sources. The Total Cost listed for the subproject likely represents the estimated amount of Center infrastructure utilized by the subproject, not direct funding provided by the NCRR grant to the subproject or subproject staff. Mass spectrometry, especially, capillary liquid chromatography-mass spectrometry (LC/MS) is the most important tool for the acceleration of knowledge acquirement of the protein machinery underpinning biomedical research. However, it is has never been successfully applied to health disparities research, which aims at determining changes in protein expression under adverse societal, behavioral or environmental conditions to identify proteins that may be involved in diseases of high-risk racial and ethnic populations. Research in health disparities requires the ability to identify and quantify proteins with a wide dynamic range in abundance, particularly for serum specimens. Information processing, comprehension and interpretation are critical. A major bottleneck of protein biomarker discovery in health disparity research by label- free, LC/MS arises from the limitations of computer algorithms that are currently available to process this data. These algorithms quantify and determine the sensitivity and specificity of putative protein with strong positive or negative correlations to a disease state. However, low abundance protein biomarkers, often the most specific, are easily missed by computer algorithms. As a result, differentially expressed candidate biomarkers are hard to identify except in biological fluids proximal to the site of disease which have been shown to be significantly enriched in proteins that derive from diseased tissue. Our preliminary study show that popular signal processing methods are inadequate for low abundance protein detection. We observed that the deficiency of current algorithms arise from 1. a lack of complete and accurate modeling of LC/MS data, and 2. suboptimal processing methods. Consequently, the dramatic new improvements in capillary LC/MS cannot yet be fully translated into optimal discovery of protein biomarkers with greater sensitivity and selectivity which are the most important for disease diagnosis and treatment. Our long term goal is to develop a suit of advanced algorithms for biomedical research including protein biomarker identification, protein quantification, function and pathway analysis using a variety of highthoughput methods, and the combination of multiple information sources such as, proteomic and gene expression data. The application of these tools will significantly strengthen the capability of UTSA in health disparity research. The objective of this proposal is to develop advanced LC/MS peak detection, alignment, feature selection, time-series data analysis and accurate protein quantification tools that are sensitive to low abundance proteins and apply the tools. Our hypothesis is that improved signal processing for LC/MS can improve the sensitivity and specificity of protein biomarker identification, quantification and functional analysis. This hypothesis will be tested by applying the tools we develop to the research project led by Dr. Forsthuber entitled, """"""""Biomarker Discovery in Glucocorticoid Resistance in EAE"""""""". To test this hypothesis and accomplish the objectives, we will carry out the following three specific aims:
Aim 1. To improve the specificity and sensitivity of protein biomarker identification using capillary LC/MS data. 1. Accurately model peptide and noise signal in capillary LC/MS datasets 2. Develop a near optimal statistical peak, picking algorithms that provide soft information based on accurate modeling. 3. Develop a peak alignment method that will resolve weak peak identities in capillary LC-MS datasets;and 4. Develop a context based feature selection algorithm for biomarker identification based on soft information provided by the peak detection and alignment algorithms.
Aim 2. To develop protein biomarker quantification and analysis tool based on time- series data. 1. Develop tools for the quantification of discovered protein markers over a time course. 2. Develop tools for analyzing time-series data for biomarker verification and analysis.
Aim 3. Verify and apply the developed algorithms Establish the quantification accuracy and detection limit of known proteins of interest such as cytokines by applying the proposed algorithm. 1. Quantify level changes of known proteins of interest in GC resistance over a time course and establish their correlation with treatments in CNS tissue, CSF and serum. Verify quantification results in step 2 with Luminex and ELISA assay. 2. Perform differential analysis in GC-resistance study to discover new potential protein biomarkers.. 3. Quantify and characterize the discovered protein marker. Beyond the scope of this current project, the developed algorithms will be applied to other researcher's sample.
Showing the most recent 10 out of 181 publications
