장연수 (영남대학교)

김주현 (영남대학교)

장연수 (영남대학교)

김민규 (영남대학교)

이솔하 (영남대학교)

임병훈 (영남대학교)

김주현 (영남대학교)

Untargeted high-resolution mass spectrometry (HRMS) enables the broad exploration of unexpected metabolites. However, the resulting complex feature profiles and in-source fragmentation often hinder systematic metabolite tracking. While Feature-Based Molecular Networking (FBMN) is widely recognized as a powerful tool for visual data analysis, researchers frequently face significant challenges in obtaining biologically meaningful networks due to the high sensitivity of network topology to parameter settings. Therefore, this study established a practical and optimized workflow to streamline FBMN-based metabolite exploration, using the metabolism of eutylone as a model pathway. To reduce network complexity, we successfully prioritized incubation-associated features by filtering out background features present in 0-min control samples. Furthermore, Ion Identity Molecular Networking was integrated to systematically link parent ions with their corresponding in-source fragments within a single network. Crucially, we demonstrated how empirical adjustments of critical parameters, specifically the minimum matched peaks and cosine similarity thresholds, directly impact network connectivity and the boundary of captured metabolite clusters. This study provides practical insights and hands-on experiences in parameter tuning, helping researchers transform chaotic untargeted HRMS datasets into highly resolvable and interpretable metabolic maps.

Acknowledgement: This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korean government (MSIT) (RS-2026-25479461).