A Metabolism-Informed Neural Network Identifies Pathways Influencing the Potency and Toxicity of Antimicrobial Combinations.

Although CALMA does not surpass existing methodologies that use Random Forests, the rationale for utilizing a mechanistic ANN architecture in CALMA is for enhanced interpretability without compromising performance, as evidenced by highly significant predictions comparable to existing methods, and flexibility to predict both toxicity and potency.

also in 132,142 other papers