Mycobacterial infections remain a major global health concern, requiring fast and accurate identification for effective treatment. Due to the species diversity and geographic heterogeneity of mycobacteria (MB), distinguishing them at the subspecies level is particularly important in clinical settings. However, existing methods, such as MALDI-TOF mass spectrometry, are limited in their ability to achieve this resolution and require a time-consuming liquid culture step, delaying critical treatment decisions.

To overcome these challenges, Prof. Fan from Tulane University and colleagues developed a peptidome-based method using mycobacterial growth indicator tube (MGIT) cultures for faster and more precise identification. By optimizing sample extraction through acetonitrile precipitation on culture filtrate protein (CFP) from MGIT cultures, they eliminated the need for lengthy liquid culturing. They then applied LC-MS/MS with a PEAKS database search to identify peptides with high specificity.

A key breakthrough in their approach is a decision tree-based algorithm that classifies mycobacteria not just at the species level, but also at the subspecies level—something that typically requires whole-genome sequencing (WGS) in conventional workflows. When tested on real-world clinical samples, the method achieved a perfect score, demonstrating its strong potential for rapid and reliable diagnosis.
Given the complexity of MB species, the researchers ensured high confidence in their identifications by manually verifying PSMs in the PEAKS database search and confirming support ions for each differential amino acid residue. With the release of PEAKS 12, the introduction of a “confident mutation” step in PEAKS SPIDER search now enables this validation process to be fully automated.

This new workflow offers a faster, more accurate, and clinically viable alternative to existing methods, improving diagnostic efficiency and patient outcomes.

 Read the full article here:
https://www.embopress.org/doi/full/10.1038/s44321-025-00207-5 (edited)