FederEI: Federated Library Matching Framework for Electron Ionization Mass Spectrum Based Compound Identification.

Abstract

Mass spectrometry (MS) is a powerful tool for compound structure elucidation, crucial in various domains including pharmaceuticals, chemical synthesis, environmental analysis, and metabolomics. Compound identification via spectral matching relies heavily on reference libraries, yet existing universal libraries may lack comprehensiveness. Laboratories often create their own libraries, leading to fragmentation within the scientific community. Addressing this, we propose FederEI, a federated library matching solution for EI-MS-based compound identification. FederEI enables decentralized compound identification while preserving data privacy, achieved through a server-to-server connection framework. Each laboratory retains control over its spectral data, enhancing privacy and minimizing data transmission. FederEI integrates a fast and accurate library matching algorithm, offering comparable performance to local matching while maintaining data security. Evaluation against the classic approach demonstrates FederEI's effectiveness, ensuring robust compound identification.