“…Moreover, the CA and IA models basically assume that all components are either similarly or dissimilarly acting compounds, respectively. The CA and IA models have been extensively reviewed and compared to each other. − If the mode of toxic action of each mixture component is not known appropriately, the CA model is preferred in mixture risk assessment over the IA model due to its simplicity (i.e., less data demanding than IA) and conservative nature (i.e., CA tends to predict higher toxicity effects than IA). ,,,, Recently, various integrated addition models combining CA with IA have been developed to consider both similarly and dissimilarly acting compounds by employing different computational toxicology methods based on machine learning algorithms and theoretically calculated features, for example, quantum chemical descriptors or molecular descriptors. − This is due to the fact that living organisms can be simultaneously exposed to both types of compounds via multiple environmental exposures. Although such integrated addition models were designed to overcome the limitations of the CA and IA models as well as to predict the mixture toxicity more accurately than both models, ,,− they basically ignored the synergistic toxicity.…”