Artificial Intelligence in Precision Dentistry: In Silico Development of DNA Aptamers Targeting Oral Cancer Biomarkers

Oral squamous cell carcinoma (OSCC) is an aggressive oral cancer in which delayed molecular diagnosis limits early detection and treatment outcomes. Artificial intelligence (AI)-enabled computational methodologies are increasingly being integrated into dentistry to accelerate biomarker discovery, molecular diagnostics, and precision healthcare. Aptamers are short, single-stranded three-dimensional (3D) nucleic acids capable of specifically binding target proteins. This study aimed to computationally engineer novel in silico DNA aptamers targeting prognostic OSCC protein biomarkers. DNA aptamers were designed and characterised against Ki-67, epidermal growth factor receptor (EGFR), mutant p53, and vimentin using computational biology and cancer bioinformatics approaches. Transfer RNA-derived sequences from Homo sapiens, Mus musculus, Escherichia coli, and Schizosaccharomyces pombe were truncated and optimized. Secondary (2D) and 3D aptamer structures were generated using Mfold and RNAComposer, followed by RNA-to-DNA conversion and structural refinement using PyMOL and AutoDock Tools. Molecular docking with AutoDock Vina evaluated aptamer–protein binding affinities, while Protein–Ligand Interaction Profiler (PLIP) characterised hydrogen-bond (H-bond) distances. Molecular dynamics simulations using GROMACS assessed conformational stability through root-mean-square deviation (RMSD) and root-mean-square fluctuation (RMSF) analyses. Computational outputs were evaluated descriptively and comparatively without inferential statistical analysis. DNA aptamers with stem and hairpin-loop conformations ranging from 25- to 50-mers demonstrated strong binding affinities of −16 to −20 kcal/mol against OSCC protein biomarkers, with favourable H-bond interactions <4 Å. RMSD analyses demonstrated stable aptamer–protein interaction equilibration between 0.01 and 0.05 nm, whereas RMSF analyses revealed localized flexibility at distinct aptamer regions essential for target recognition. These findings highlight the potential of AI-enabled computational biology and molecular modelling to accelerate biomarker-targeted molecular probe development and support next-generation diagnostic platforms for oral cancer and other dental diseases.