Preclinical disease models play an important role in the establishment of new treatment paradigms, identification of biomarkers and assessment of drug efficacy and safety. However, the accuracy of these models in context of the human disease are sometimes questioned, e.g. due to trials failing to confirm efficacy in humans. We suggest that one reason behind this gap in predictability may relate to how the preclinical data is analyzed and interpreted. In the present paper, we introduce a holistic approach to analyze and illustrate data in context of one of the most commonly used colitis models, i.e. the mouse dextran sulphate sodium (DSS) colitis model. Diseased mice were followed over time along disease progression and by use of tool pharmacological compounds activating nuclear hormone receptors, respectively. A new multivariate statistics approach was applied including principal component analysis (PCA) with treatment prediction subsequent to establishing the principal component analysis model. Thus, several studies could be overlaid and compared to each other in a new, comprehensive and holistic way. This method, named mouse colitis global property screening, appears applicable not only to any animal modelling series of studies but also to human clinical studies. The prerequisites for the study set up and calculations are delineated and examples of new learnings from the global property screening will be presented.