Unlocking predictive genetic factors with artificial intelligence: relationship between dental impaction and hypodontia evaluated via association-rule algorithms: a case-control study.

Artificial intelligence (AI)-driven genetic analysis is crucial for the advancement of personalized dental treatments. Despite progress in genetic research, its clinical application remains limited. This case-control study utilized an association-rule mining algorithm (Apriori) implemented using Python software (version 0.22.0, 2024) to predict dental impaction based on MSX1, PAX9, and AXIN2 polymorphisms. The study was carried out at two centers in Saudi Arabia in October 2023 and involved 106 participants (42 males, 64 females; mean ± standard deviation age = 20.56 ± 8.07 years). Participants were categorized into 52 controls and 54 individuals with hypodontia, among whom 13 had impacted teeth. Saliva samples were analyzed for three single nucleotide polymorphisms: AXIN2 (rs2240308), PAX9 (rs61754301), and MSX1 (rs12532). Although multinomial logistic regression analysis indicated genotypic variations, no statistically significant associations with dental impaction were identified (P = 0.112). However, association-rule mining identified notable genotype patterns with the MSX1 A/A genotype (support = 0.224, confidence = 0.827, lift = 1.475). The combination of PAX9 (C/C) and MSX1 (A/A) had the highest predictive value (lift = 1.671), followed by MSX1 (A/A) with AXIN2 (G/G) (lift = 1.646), and PAX9 (C/C) with AXIN2 (G/G) (lift = 1.475). Based on available scholarly literature, this is among the pioneering studies to use association-rule algorithms to predict dental impaction, highlighting the potential of AI in genetics-based orthodontic diagnostics.