Methods for proteomic diagnosis and prognosis in inflammatory diseases
Research Objectives:
Clinical proteomics offers new opportunities and challenges in the construction of prediction methods for diagnosis and prognosis. A special feature of proteomics is the complex nature of the spectral proteomic signal which is measured, as it consists of high-dimensional functions which represent the within-patient proteome expression. This project will develop new predictive diagnostic and prognostic methodology for mass spectrometry based clinical proteomics. We consider several motivating examples and datasets in oncology (breast, colon and pancreatic cancer). The project is an interdisciplinary collaboration with the Departments of Surgery (Dr. W. Mesker and Prof. Dr. R. Tollenaar) and the Biomolecular Mass Spectrometry Unit (Dr. Y. van der Burgt and Prof. Dr. A. Deelder).
Description of work:
We consider several related research questions within this project.
-
How can several repeat proteomic measures within the same patient be combined to obtain improved predictions?
-
How can distinct omics measures (e.g. genomic and proteomic) within the same patient be combined for predictive purposes?
-
How can calibration of prediction be assured within any single omics source predictor?
-
How can relative added value of distinct omics sources be evaluated within the predictive paradigm?
Host Institution: Leiden University Medical Center