Machine learning risks errors in enzyme function prediction without careful tuning
The Accuracy Paradox: Empirical Diagnostic of Default Decision Thresholds in Multi-Label Enzyme Commission Prediction [With Code]
Machine LearningArtificial Intelligence
Summary
Predicting enzyme functions automatically helps scientists understand biology and design drugs, but usual machine learning methods often assume simple rules that don’t fit rare cases well. The authors found that using standard prediction thresholds makes the system look accurate overall, but it actually misses many important enzyme types. Some enzyme classes are predicted too much while others are barely recognized, causing big mistakes. They suggest special tuning for each enzyme type and calibration steps to avoid these hidden errors and make predictions more trustworthy.
Enzyme Commission numbersmulti-label classificationdecision thresholdclass imbalanceaccuracy paradoxmacro F1-scorerecallROC-AUCpost-hoc calibrationmachine learning
Authors
Bilal Ahmad, Rajed Mehmood
Abstract
Automated prediction of Enzyme Commission (EC) numbers plays a central role in functional annotation and computational drug discovery. However, standard multi-label machine learning pipelines frequently rely on default decision thresholds (t=0.50), assuming balanced prior distributions across target heads. In this study, we present a systematic empirical diagnostic of uncalibrated fixed decision boundaries operating under severe class imbalance across N = 14,096 annotated compounds categorized into six primary EC classes (EC1-EC6). Our results highlight a pronounced Accuracy Paradox: while the multi-label system achieves a deceivingly high mean accuracy of 77.16%, the macro F1-score (0.3976) and macro recall (0.3872) reveal severe predictive breakdown. Majority target classes suffer from hyper-sensitivity and over-prediction, whereas minority classes exhibit sharp recall decay, culminating in a total decision boundary collapse for EC6 (Recall = 0.00%) despite underlying discriminative power (ROC-AUC = 0.5857). Feature correlation analysis further reveals high linear redundancy among topological indices relative to fingerprint density metrics. Ultimately, this diagnostic study demonstrates that standard point predictions mask critical errors in bioinformatics workflows. We establish target-specific threshold optimization and post-hoc conformal calibration as essential, open-source post-processing safeguards for reliable applied machine learning and deep learning architectures.