AI-driven, qualitative GC–MS/MS-based endogenous steroid profiling with automated classification for sex cohort determination

The proper identification of sex cohorts is essential for anti-doping within the regulatory framework for horse racing, as investigations could be compromised if the sex of a stallion, mare or gelding were misclassified. The single-threshold approach is not very useful for assessing hormonal profiles in horses because of nonlinear relationships among testosterone and other hormones and metabolites. In the present study, an automated Steroid Detection System (SDS) was developed using PDF-based data extraction and characteristic qualitative metabolic profiles and peak intensity patterns obtained by GC–MS and utilized in AI driven model. A total of 291 post-race urine samples from Males ( n = 122), Females ( n = 102), and Geldings ( n = 67) collected at Seoul Racecourse, Korea, were analyzed by GC–MS/MS in multiple reaction monitoring MRM mode. IS-normalized ratios of 10 key biomarker compounds were used as input features. The Random Forest (300 decision trees, balanced class weighting) had a 5-fold stratified cross-validation accuracy of 90.7 ± 2.3% with per-class accuracies of 98.4% (Male), 87.3% (Female) and 82.1% (Gelding). On a stratified held-out test set ( n = 45), it achieved 95.3% accuracy, F1-score of 0.95, and AUC of 0.98, outperforming Logistic Regression (81.4%, 0.80, 0.85), Decision Tree (84.7%, 0.84, 0.88), SVM (87.2%, 0.87, 0.92), and Gradient Boosting (93.1%, 0.93, 0.97). The feature importance analysis showed that 5α-androstane-3β,17α-diol (EN3), testosterone (EN1) and nandrolone (EN4) were the most important classifiers with 61.2% of the signal. Multivariate steroid profiling using IS normalization and Random Forest is a powerful, automated tool for gender determination in horses for regulatory anti-doping control.