Confidence-MS: Confidence-based identification and quantification of pathogenic taxa and protein functions in time-critical diagnostic scenarios using tandem mass spectrometry

• Applicant: Privatdozent Dr. Thilo Muth
• Subject Area: Bioinformatics and Theoretical Biology
• Term: from 2020 to 2025
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 439719438

Pathogens of bacterial or viral origin pose a significant threat to public health worldwide, as demonstrated by the recent SARS-CoV-2 pandemic. This highlights the need for rapid and precise diagnostic techniques during outbreaks to reduce the time to diagnosis and treatment, thereby mitigating the risks of virus transmission and mortality. In addition to conventional methods such as PCR, open approaches like next-generation sequencing (NGS) and mass spectrometry (MS) have become particularly valuable, with MS distinguished by its high accuracy and reproducibility in detecting virus-specific proteins. The Confidence-MS project aims to explore the transformative potential of MS-based proteomics to improve pathogen detection and understanding. Building on the developments of the ViPPip project, which developed statistical and bioinformatic methods for detecting viral pathogens, the focus is on data analysis in time-critical situations, especially during viral outbreaks. This includes distinguishing viral subtypes in novel infection sources, both in known viral infections and newly emerging wild-type virus outbreaks. Addressing the challenges in these scenarios, such as the detection of strain-specific peptides and robust taxon quantification, requires an update to the latest MS technologies and a more pragmatic approach to applying our methodology in diagnostics. The methodology of the Confidence-MS project aims to increase the accuracy and reliability of pathogen detection using proteomic MS technology. First, we aim to extend the probabilistic model developed in the initial funding period to enable robust, confidence-based quantification of pathogenic taxa. Second, we will develop new approaches based on this model to fully cover the functional spectrum of proteins derived from pathogen samples. Third, we plan to provide a tailored search strategy for taxonomic and functional inferences, accommodating both DDA and novel DIA MS technologies. Fourth, we will implement user-friendly modules for applying existing methods and establish a workflow for simulating realistic benchmarking data with taxon abundance profiles. Finally, we aim to formulate best practices and recommendations within our framework for taxon classifications in diagnostic scenarios, such as outbreaks of novel viruses.

• DFG Programme: Research Grants