Abstract- High-throughput sequencing has generated over 246 million protein sequences, yet only a fraction have been experimental validated. This annotation gap is particularly severe in plants, where genome plasticity and frequent gene duplication mean that proteins with >70% sequence identity can have completely different functions. Plants produce 200,000 to 1,000,000 specialised metabolites through this diverse enzymatic machinery, but traditional homology-based predictions fail to capture this functional diversity. These observations suggest the classical sequence-structure-function paradigm requires additional dimensions to accurately predict enzyme activity in plants.

My PhD developed a novel approach to improve functional classification of plant enzymes by comparing 3D active site electrostatic maps across enzyme families, proposing that electrostatic properties represent a critical intermediate between structure and function.

I first evaluated established methods including sequence similarity networks, evolutionary velocity, and phylogenetic analysis to understand their limitations in plant systems. I then developed a computational framework to generate and compare pairwise electrostatic field maps for over 500 homologous enzymes. Since enzyme function fundamentally depends on the electrostatic environment that guides substrate binding and catalysis, this approach captures functional information that sequence and structure alone cannot provide.

Analysis of electrostatic signatures successfully distinguished enzymes by taxonomic origin, predicted pH optima, and identified subcellular localisation patterns. This electrostatic mapping approach opens new avenues for predicting enzyme function, particularly in poorly characterised plant systems, understanding enzyme evolution, and guiding rational enzyme design for synthetic biology applications.