Transfer learning with attention-enhanced deep neural networks for plant enzyme classification using UniProt protein embeddings

Accurate classification of enzymes and non-enzymes from protein sequences is fundamental to understanding plant metabolism, with applications in crop improvement and biotechnology. While deep learning has shown promise for protein function prediction, most approaches train from raw sequences requiring extensive computational resources. Transfer learning using pre-trained protein embeddings offers an efficient alternative, yet its application to plant enzyme classification across multiple species remains unexplored. We present a transfer learning framework combining UniProt-derived protein embeddings with attention-enhanced and baseline deep neural networks for enzyme classification in four major plant species: Arabidopsis thaliana , Brassica species, Oryza sativa , and Triticum aestivum . Our curated dataset comprises 22,267 unique protein sequences with 1024-dimensional embeddings. To prevent data leakage, we implemented homology-aware splitting using CD-HIT clustering at 60% identity, creating 13,784 clusters. We also performed Leave-One-Species-Out evaluation to assess cross-species generalization. Systematic benchmarking of 10 models reveals that the Attention-Enhanced DNN achieves the most consistent performance, with test AUC of 0.987 ± 0.001, accuracy of 0.955 ± 0.004, F1-score of 0.939 ± 0.005, and MCC of 0.903 ± 0.008, while demonstrating robust cross-species generalization with average LOSO AUC of 0.985 ± 0.008. Statistical testing confirmed significant differences among methods (Friedman test, p < 0.001), with the proposed framework winning 67% of hyperparameter configurations. This study demonstrates that pre-trained UniProt embeddings combined with deep learning provide an effective approach for plant enzyme classification, with the attention mechanism contributing modest yet statistically significant improvements.