Machine Learning‐Driven Screening of Promising Mutants in Nilgirianthus ciliatus Callus Cultures for Enhanced Phytochemicals Production

The unsustainable exploitation of Nilgirianthus ciliatus for its pharmaceutically important bioactive compounds has led to its endangered status, necessitating innovative strategies for its conservation and utilization. This study establishes a scalable in vitro platform for enhanced biomass and phytochemical production through induced mutagenesis and AI‐driven predictive modeling. Callus cultures were subjected to individual and combinatorial chemical mutagenesis using ethyl methanesulfonate (EMS) and sodium azide (SA) under optimized treatment regimes. Callus cultures were evaluated for enzymatic and non‐enzymatic antioxidant responses using standard biochemical assays, while squalene accumulation was quantified via high‐performance thin‐layer chromatography (HPTLC). Machine learning (ML) models: Multilayer Perceptron (MLP), Random Forest (RF), and Light Gradient Boosting Machine (LightGBM), were trained on experimental datasets to predict key growth and metabolite parameters. The optimized treatment (0.05% EMS + 0.05% SA, 30 min) led to a 3.76‐fold increase in squalene yield (308.39 μg mg −1 ), with concurrent improvements in biomass and antioxidant indices. MLP demonstrated the highest predictive accuracy ( R 2 = 0.971), validating its application for forecasting complex biological outcomes. This integrated framework not only offers a scalable in vitro strategy for sustainable metabolite production but also minimizes pressure on wild populations. The framework offers strong translational potential for industrial‐scale metabolite production and serves as a model for data‐driven optimization in medicinal plant biotechnology.