Restoring comparative expertise to improve translation of animal models

Animal models remain indispensable in biomedical research, particularly for discovering disease mechanisms within intact living organisms. Yet confidence in their human translational relevance has been weakened by apparent failures of preclinical findings to predict clinical outcomes. Although often attributed to species differences, such failures also reflect correctable problems in model selection, experimental design, endpoint choice and phenotype interpretation. A key, but overlooked, contributor is the erosion of comparative expertise in pathobiology and clinical medicine needed to relate animal phenotypes to human disease across species, strain, and developmental, physiological and pathological contexts. This expertise gap can lead investigators to mismatch models to biological questions, misinterpret background abnormalities as engineered phenotypes or extend mechanism-focused models beyond the questions they are suited to answer. New approach methodologies (NAMs), such as in vitro and in silico models, provide important complementary tools, but do not eliminate the need for whole-organism models or resolve this interpretive problem. Ensuring the reliability of biomedical research will require preserving well-justified animal models where whole-organism systems biology remains necessary, integrating validated NAMs only where they are fit for purpose, and rebuilding the comparative expertise needed to interpret both model types rigorously.