Résumé
Integrating multi-omics data is a comprehensive method for under - standing cellular or organismal responses to chemical exposure. In Canzler et al. (2020) [1], we described best practices for conducting multi-omics studies. Here, we present a multi-omics investigation following these best practices integrating clinical, histopathological, and six layers of omics data (long and short transcriptomics, proteomics, tissue and plasma metabolomics, and phosphoproteomics). Utilizing the well-studied compounds Phenytoin and Propylthiouracil (PTU) in a rat toxicity study over 28 days with an additional 14-day recovery period, we aimed to explore mechanisms of direct and indirect thyroid toxicity.Our findings demonstrate that multi-omics approaches significantly surpass single-omics analyses in identifying regulatory pathways and molecular effects relevant to toxicology. For instance, the multi-modal data elucidated complex responses to PTU and Phenytoin, both at the transcript and protein levels and in metabolomic shifts, offering insights into the perturbations of thyroid hormone biosynthesis and liver metabolic pathways, respectively. Also, we found the combined interpretation of omics-, clinical, and histopathological parameters particularly beneficial. Importantly, the simultaneous data integration reveals intricate interplays between different omics layers, highlighting how individual and combined data layers uniquely contribute to understanding toxicological outcomes. Furthermore, grouping approaches focusing on common molecular effects benefit substantially from multiple omics layers.This study emphasizes the superiority of multi-omics in detecting molecular responses to toxicants, deepening our understanding of mechanisms of action and enhancing the predictive capabilities of toxicological assessments. Our work emphasizes the value of integrated multi-omics strategies in advancing the field of toxicology towards more holistic and mechanistically informative evaluations, potentially informing regulatory decision-making and risk assessment processes.