Organophosphorus compounds (OPCs) are widely applied in diverse industries, yet their continuous release across the full life cycle poses significant risks to human health and ecosystems. Current research on OPCs is largely limited to organophosphate esters (OPEs), with scarce attention paid to structurally diverse OPCs─creating critical gaps in comprehensive environmental screening and risk assessment. To address this gap and facilitate robust OPC analysis, we curated and integrated OPC-related cheminformatics data from academic and public sources, constructing a structured library of OPCs classified by chemical structure. We further prioritized high-production-volume (HPV) OPCs, which are pivotal for targeted environmental research and regulatory focus. A key analytical advancement herein is the development of a lightweight deep learning model based on graph neural networks (GNNs), which extracts molecular structural features to predict tandem mass spectra (MS/MS) data for OPCs. This model directly overcomes the major bottleneck of limited reference MS/MS for most OPCs in existing libraries. These curated resources and predictive capabilities are embedded into DCOP (Database of Chemicals for Organophosphorus)─a high-quality, large-scale, open-access software platform integrating search query, MS/MS prediction, and data download functions. DCOP fills unmet needs in OPC research, providing a valuable tool to advance environmental monitoring, risk assessment, and regulatory decision-making for OPCs globally.
Bi et al. (Fri,) studied this question.