Project

A computational pipeline for highly sensitive profiling of tissue leakage proteins in liquid biopsies

Code
3S007320
Duration
01 November 2020 → 31 October 2024
Funding
Research Foundation - Flanders (FWO)
Research disciplines
  • Medical and health sciences
    • Medical proteomics
    • Development of bioinformatics software, tools and databases
    • Bio-informatics and computational biology not elsewhere classified
    • Protein diagnostics
    • Biomarker discovery and evaluation not elsewhere classified
Keywords
Clinical proteomics Bioinformatics Machine learning tissue prediction
 
Project description

A biomarker is a factor found in the body that indicates the presence of disease, ideally at an early stage. The detection of biomarkers in liquid biopsies such as blood is becoming increasingly important in a clinical context. Often, disease-specific proteins in liquid biopsies are tissue leakage proteins (TLPs), derived from tissues directly affected by the disease and thus early signs of this disease. However, such TLPs tend to occur at low concentrations, requiring sensitive approaches. One of the most sensitive means to detect proteins is mass spectrometry, and the most recent development in this field is data-independent acquisition (DIA) technology. Yet, even though DIA is the most popular protein biomarker discovery technology in clinical practice, there are still opportunities to improve the sensitivity of detection with more advanced bioinformatics and statistics approaches. This is especially important for the reliable detection of low abundance TLPs, which will benefit from a tissue-specific protein expression catalogue. This because it currently is difficult to allocate detected protein markers to their tissue(s) of origin. I will here therefore develop the bioinformatics components needed to enable sensitive DIA for TLPs. I will moreover integrate these components into a complete workflow, which I will then validate on both public data, as well as on relevant in-house data from clinical samples in neurodegenerative disease and cancer contexts.