Label-free quantification of analytes is gaining recognition as a very good strategy for biomarker discovery. However, such quantification is typically not addressed adequately in the instrument specific software packages. Biological variability and disease heterogeneity in human populations further complicate the MS-based biomarker discovery. The specific research aims of this project are the following: (1) Investigate data preprocessing algorithms for label-free quantification of peptides in serum using liquid chromatography?mass spectrometry (LC-MS) technologies. The initial focus will be on investigating and developing normalization, alignment, and peak detection methods to preprocess LC-MS spectra generated by two LC-MS instruments. (2) Investigate feature selection algorithms that can isolate subgroup-specific biomarkers; this process will account for biological variability and disease heterogeneity among human populations.
Spike-in study will be conducted to obtain replicate LC-MS spectra with known peptide differences. The spectra from this study will be used to optimize the proposed data preprocessing and feature selection algorithms, and compare them with other existing solutions. The optimized analytical tools will be applied to select the most useful peaks for detecting hepatocellular carcinoma (HCC) at a treatable stage. Blood samples collected from patients with cirrhosis, HCC cases, and healthy controls in Egypt, United States, and Thailand will be used. The peptides that correspond to the selected peaks will be identified by sequencing the peaks using MS/MS instruments and complementary methods. The ability of these candidate peptide markers to detect HCC will be validated using appropriate isotope dilution mass spectrometric assays.
Novel aspects of the proposed biomarker discovery algorithms include the capability to (1) normalize and align spectra to allow the discovery of candidate peptide biomarkers whose performances are not influenced by non-disease related artifacts in the data, (2) detect peaks that are consistent with the ones manually selected by MS experts, (3) select biomarkers through an approach that accounts for biological variability and disease heterogeneity among subjects, and (4) provide a framework for a complete analytical pipeline beginning from sample preparation to validation of candidate peptide biomarkers.
The project will establish new interdisciplinary research collaboration in the areas of bioinformatics, mass spectrometry, and tumor biology, in which faculty and student members of the project will be cross-trained to transcend traditional disciplinary boundaries. It will provide research-enriched learning experiences to students through two educational activities at Georgetown University Medical Center.
Further information on the project may be found at the project web site: http://microarray.georgetown.edu/ressomlab
