QSAR (Quantitative Structure-Activity Relationships) model constructed based on comprehensive toxicity action mode classification for predicting acute toxicity of organic compound to daphnia magna

Abstract

The invention discloses a QSAR (Quantitative Structure-Activity Relationships) model constructed based on comprehensive toxicity action mode classification for predicting acute toxicity of an organic compound to a daphnia magna. At first, a comprehensive toxicity action mode classification method is developed, and the method is adopted to classify the compound based on structure characteristics. After the compound is classified according to the method, an EC50 model is constructed on the basis of a molecular structure descriptor of the compound through a multiple linear regression method, so that an EC50 value of acute effective toxicity of the organic compound to the daphnia magna can be quickly and effectively predicted; and the goodness of fit, the robustness and the predictive ability of the model are represented according to construction and verification guidelines of an organization for economic cooperation and development about the QSAR model; the model application domain is clear; the prediction model is simple and easy to be programmed, and an important data support is provided for acute toxicity assessment of organic chemicals.

Description

technical field [0001] The invention covers a classification method for comprehensive toxicity action modes, and a method for predicting the acute toxicity of organic compounds to Daphnia magna by establishing a quantitative structure-activity relationship model (QSAR) based on the classification method, and belongs to the field of ecological risk assessment testing strategies. Background technique [0002] The REACH regulation (Registration, Evaluation, Authorization and Restriction of Chemicals) issued by the European Union requires that all chemicals entering the EU market need to be subject to preventive management, and plans to conduct toxicity assessment on chemicals entering the EU market exceeding 1 ton per year. However, this task faces huge challenges. On the one hand, the current chemical information is very lacking. For example, according to the list of commercialized chemicals in the EU, the total number of existing chemicals and new chemicals has reached 132,119...

AI Technical Summary

Problems solved by technology

Although the model constructed based on the machine learning method has a good degree of fitting, the model algorithm and form are not transparent enough, and the mechanism is poorly interpretable, which will affect the credibility and acceptability of the results.

At present, the models constructed based on MOA classification are mainly aimed at anesthetic compounds, and there are few studies on reactive compounds, mainly due to the complexity of MOA classification, which limits the development of the model.

Images