Artificial Intelligence Assisted Optimization of Ramaria obtusissima Extracts and Their Integrated Chemical and Biological Characterization

In this study, the biological activities of extracts obtained from <i>Ramaria obtusissima</i> were optimized using response surface methodology (RSM) and artificial neural networks-genetic algorithm (ANN-GA) approaches, and the chemical and biological profiles of the obtained extracts were evaluated with a holistic approach. Antioxidant potential was determined using FRAP, DPPH, TAS, TOS, and OSI parameters. It was found that the extract optimized with ANN-GA had significantly higher FRAP (242 ± 3 mg Trolox equivalent/g), TAS (6.64 ± 0.04 mmol/L), and DPPH (154 ± 3 mg Trolox equivalent/g) values compared to the RSM extract, while its OSI value was lower. Anticholinesterase activities were evaluated using IC₅₀ values, and it was determined that the ANN-GA extract exhibited a stronger inhibitory effect on acetylcholinesterase (95 ± 2 µg/mL) and butyrylcholinesterase (125 ± 3 µg/mL) compared to the RSM extract. Antiproliferative effects were investigated in A549, MCF-7, and DU-145 cell lines, and a significant and dose-dependent suppression of cell proliferation was observed in all three cell lines, particularly at concentrations of 100 and 200 µg/mL. The chemical profile was determined using LC-MS/MS and GC-MS techniques. Higher levels of phenolic compounds such as gallic acid (6694.5 ± 4.9 mg/kg), caffeic acid (3374.8 ± 4.9 mg/kg), and quercetin (1563.1 ± 2.3 mg/kg) were found in the ANN-GA extract. GC-MS analyses showed that the ANN-GA extract has a richer lipophilic component profile in terms of biologically active fatty acids and ester derivatives. The findings reveal that AI-assisted optimization offers a powerful and effective approach to enhancing the biological efficacy of mushroom-derived natural products.