Project

Advanced AI-driven Bioinformatics Solutions for Performant Peptidomics

Code
G010023N
Duration
01 January 2023 → 31 December 2026
Funding
Research Foundation - Flanders (FWO)
Promotor-spokesperson
Research disciplines
  • Natural sciences
    • Computational biomodelling and machine learning
    • Development of bioinformatics software, tools and databases
    • Structural bioinformatics and computational proteomics
    • Proteomics
Keywords
Peptidomics Deep learning Mass spectrometry Bioinformatics proteomics
 
Project description

Peptidomics is not a new field, but is currently rapidly expanding. The promise of looking at peptide components of tissues or organisms has been well-known for a long time, but methodological issues have held the field back. The renewed interest today is primarily driven by better wet-lab approaches and improved instrumentation, but the necessary bioinformatics algorithms to interpret the acquired data currently lag behind, creating a serious bottleneck for this highly promising field. We here therefore propose to address this issue by leveraging our world-leading, advanced artificial intelligence and machine learning driven approaches to build a vastly improved suite of tools to identify and quantify peptidomics data at unprecedented senstivity and specificity of identification, and at as-yet unseen quantitative accuracy and precision. We will tackle this overall challenge in four goals: (i) to create optimally performant predictors of truly generic peptide sequences' behaviour in LC-MS/MS analyses; (ii) to build an open modification peptidomics search engine that leverages these predictive models for identification of peptidomics data at unprecedented depth; (iii) to build a bespoke quantification and differential analysis pipeline for these peptidomics data; and (iv) to apply the resulting end-to-end peptidomics pipeline to several relevant and diverse peptidomics data sets as key proof-of-concepts.