MOTIVATION: Microbial biotransformation plays a central role in the environmental degradation of chemical contaminants, driven by the catalytic activities of diverse enzymes. However, linking specific enzymes to contaminant removal and predicting associated transformation products (TPs) under real-world conditions remain a major challenge. In this study, we present a self-supervised, contrastive fine-tuning strategy for reaction fingerprint learning, designed to improve the chemical relevance of BERT-based reaction embeddings for environmental biotransformation reactions. Specifically, we fine-tuned a BERT encoder such that the cosine similarity between its reaction fingerprints aligns with the Tanimoto similarity of traditional structure-based fingerprints. RESULTS: The resulting compact, 256-dimensional fingerprints, termed crxnfp, showed an improved ability to cluster reactions according to transformation type and focus attention on chemically meaningful reaction centers. Our crxnfp fingerprints were further validated in reaction classification tasks across multiple datasets, achieving superior or comparable performance relative to existing methods. Importantly, they enabled a similarity-based association of biotransformation rules and reactions from enviPath with enzyme annotations from the Rhea and UniProt databases, offering a scalable approach to enrich environmental biotransformation datasets with enzymatic information. Additionally, crxnfp was employed to identify specific enzyme classes involved in contaminant biotransformation, which were subsequently validated through experiments conducted in this study, achieving 91.3% accuracy at the third-level enzyme classification. The crxnfp fingerprints offer a promising solution to advance the understanding of contaminant biotransformation and guide the development of enzyme-informed strategies for contaminant management across diverse environmental contexts. AVAILABILITY AND IMPLEMENTATION: Code is available at https://github.com/zhangky12/crxnfp and https://github.com/zhangky12/crxnfp_knn.

MOTIVATION: Transformation products (TPs) of man-made chemicals, formed through microbially mediated transformation in the environment, can have serious adverse environmental effects, yet the analytical identification of TPs is challenging. Rule-based prediction tools are successful in predicting TPs, especially in environmental chemistry applications that typically have to rely on small datasets, by imparting the existing knowledge on enzyme-mediated biotransformation reactions. However, the rules extracted from biotransformation reaction databases usually face the issue of being over/under-generalized and are not flexible to be updated with new reactions. RESULTS: We developed an automatic rule extraction tool called enviRule. It clusters biotransformation reactions into different groups based on the similarities of reaction fingerprints, and then automatically extracts and generalizes rules for each reaction group in SMARTS format. It optimizes the genericity of automatic rules against the downstream TP prediction task. Models trained with automatic rules outperformed the models trained with manually curated rules by 30% in the area under curve (AUC) scores. Moreover, automatic rules can be easily updated with new reactions, highlighting enviRule's strengths for both automatic extraction of optimized reactions rules and automated updating thereof. AVAILABILITY AND IMPLEMENTATION: enviRule code is freely available at https://github.com/zhangky12/enviRule.

MOTIVATION: Current methods for the prediction of biodegradation products and pathways of organic environmental pollutants either do not take into account domain knowledge or do not provide probability estimates. In this article, we propose a hybrid knowledge- and machine learning-based approach to overcome these limitations in the context of the University of Minnesota Pathway Prediction System (UM-PPS). The proposed solution performs relative reasoning in a machine learning framework, and obtains one probability estimate for each biotransformation rule of the system. As the application of a rule then depends on a threshold for the probability estimate, the trade-off between recall (sensitivity) and precision (selectivity) can be addressed and leveraged in practice. RESULTS: Results from leave-one-out cross-validation show that a recall and precision of approximately 0.8 can be achieved for a subset of 13 transformation rules. Therefore, it is possible to optimize precision without compromising recall. We are currently integrating the results into an experimental version of the UM-PPS server. AVAILABILITY: The program is freely available on the web at http://wwwkramer.in.tum.de/research/applications/biodegradation/data. CONTACT: [email protected].

MOTIVATION: The University of Minnesota Pathway Prediction System (UM-PPS) is a rule-based expert system to predict plausible biodegradation pathways for organic compounds. However, iterative application of these rules to generate biodegradation pathways leads to combinatorial explosion. We use data from known biotransformation pathways to rationally determine biotransformation priorities (relative reasoning rules) to limit this explosion. RESULTS: A total of 112 relative reasoning rules were identified and implemented. In one prediction step, i.e. as per one generation predicted, the use of relative reasoning decreases the predicted biotransformations by over 25% for 50 compounds used to generate the rules and by about 15% for an external validation set of 47 xenobiotics, including pesticides, biocides and pharmaceuticals. The percentage of correctly predicted, experimentally known products remains at 75% when relative reasoning is used. The set of relative reasoning rules identified, therefore, effectively reduces the number of predicted transformation products without compromising the quality of the predictions. AVAILABILITY: The UM-PPS server is freely available on the web to all users at the time of submission of this manuscript and will be available following publication at http://umbbd.msi.umn.edu/predict/. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.