An iterative and automated computational pipeline for untargeted strain-level identification using MS/MS spectra from pathogenic samples

Mathias Kuhring; Joerg Doellinger; Andreas Nitsche; Thilo Muth; Bernhard Y. Renard

Untargeted strain level identification via MS/MS spectra is a challenging task with respect to the excessive quantity of strains that need to be considered competitively. For solving this problem, we present TaxIt, an iterative approach that first focuses on species identification and thus limits strain identification to concise selected target databases. In our method, both iteration steps take advantage of publicly available data from the NCBI Taxonomy and Protein databases. Thereby, TaxIt enables final strain identification from recent and relevant sequence data, without the need of heavily curated, tailored or taxonomically constrained databases. TaxIt supports any MS/MS peptide data suitable for classic database search methods. With further ongoing developments, our tool might enable a wide range of applications in public health, research and potentially also clinics in the future: based on further validation, TaxIt might support diagnosis and treatment of pathogen-based infectious diseases. TaxIt is available for download under open-source license at https://gitlab.com/rki_bioinformatics. All rights reserved. No reuse allowed without permission.