The goal of this model is to predict with high probability (>99,9%) that a chemical does NOT meet the Endocrine Disruption (ED) criteria according to EU No 528/2012 and 1107/2009 regulations. This model has been designed as a front-line tool which examines only the direct mechanisms for EATS: Estrogen, Androgen and Thyroid hormones receptors binding, Steroidogenic disruption and other mechanisms altering the production, distribution and metabolism of hormones.
In this model, we defined the positive result to be “the substance is expected to have low potential for endocrine disruption”. The first step is almost finalised with promising results on ER & AR binding. Preliminary results of 2D SAR for EDCs. Our objective: reduce false positives to <<1%.
Our method is a tiered approach with 3 steps: first, we use a 2D structure-activity relationship to identify EDCs, second, we refine the results with a 3D-modelling step to assess receptor docking ability, and third, we run off-the-shelf tools to provide further evidence of endocrine disruption.
If all 3 steps don’t give any ED alert, the substance can be labelled as a substance of low ED concern" (low risk of having one of the well-known ED modes of action). At this level, it should be possible to stop there without running a consequential set of laboratory studies.
This set of three methods just takes a few hours to run and so can be performed for a very reasonable cost. If inconsistent results are determined within the three steps, further elucidation should be considered which may include laboratory studies and read-across analysis.
We expect to roll out the full battery of models this summer. Stay updated to be among the first ones to benefit from this new tool!