Hierarchical Model Prediction of Antibiotic Resistant Neisseria Gonorrhoeae

Abstract

The CDC has classified Neisseria gonorrhoeae as a significant threat with an estimated 1 million new cases per year (CDC, 2021). Due to the emergence of antimicrobial resistance (AMR) mutations, the CDC must keep shifting its recommended antibiotics to combat the sexually transmitted disease, resulting in many healthcare professionals ineffectively prescribing antibiotics to patients. This study investigates mutations within DNA contigs in order to construct a computational model to identify potential biomarkers with significant correlation for antibiotic resistance. The machine learning model Support Vector Machine (SVM) was leveraged to achieve this goal. The SVM model was trained on a dataset of 9967 samples of individual patient DNA contigs with AMR to the top 3 most widely prescribed antibiotics globally: azithromycin, ciprofloxacin and cefixime. The SVM model achieved an overall accuracy of 90.3%. This study demonstrates the potential of machine learning techniques in genome based detection methods in the translational medical field.