The invention discloses an aquatic
organism toxicity effect threshold value calculation method based on
machine learning. The method comprises the following steps: step 1, constructing a chemical target
toxicity endpoint feature
database; 2, based on the
database established in the step 1, selecting chemical molecule descriptors serving as input by adopting a
feature screening method, and constructing a target
toxicity endpoint prediction model; step 3, constructing a chemical toxic action mode
database; 4, based on the database established in the step 3, screening chemical molecule descriptors serving as input by adopting a
principal component analysis method, and constructing a toxic action mode classification model; step 5, based on the target toxicity endpoint prediction model and the toxic action mode prediction model, screening a reference compound of a target toxicity test type in the chemical mixture; and step 6, based on the reference compound, calculating the toxicity effect threshold value of the aquatic
organism target toxicity end point. According to the method, a
machine learning technology is adopted, based on a toxicological
data set in a public database, derivation of the toxicity threshold value of the
aquatic organisms is achieved, time and economic cost are saved, and a method is provided for
water quality risk assessment and management and control.