AI-assisted identification of innovative phytochemicals from Aizoon canariense aimed at brachyury protein in chordoma: a computational strategy

Chordoma is an unique and aggressive bone malignancy along with limited therapeutic options, largely due to the undruggable nature of the TBXT oncoprotein. In this study, we employed an AI-assisted drug discovery approach to optimize β-sitosterol from Aizoon canariense against TBXT. Among the generated analogs, Artificial Intelligence Ligand 2 demonstrated strong inhibitory potential, showing a stable binding free energy of - 81.09 kcal/mol and maintaining conformational stability during a 500 ns molecular dynamics simulation. Key pharmacokinetic parameters revealed blood-brain barrier permeability, high gastrointestinal absorption as well as compliance with Lipinski's rules, while toxicity evaluation predicted moderate acute toxicity (LD₅₀ = 800 mg/kg, Class IV) with no major systemic risks. Density Functional Theory analysis confirmed molecular stability and reactivity suitable for biological interactions. Overall, these results highlight AI-Ligand 2 as a promising phytochemical derivative for targeting TBXT in chordoma, providing a foundation for future experimental validation and preclinical studies.