Method for predicting cytotoxicity of disinfection by-products

• Main Authors: TANG MENGMENG, ZHOU QING, HAN LIANGLIANG, SHUANGDONG DONG, PAN YANG, SHI PENG, WANG LEYI, LI AIMIN, REN JIAFENG, CHEN XUEYAO, LUO JIAYI
• Format: Patent
• Language: Chinese; English
• Published: 30.08.2022
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTEDFOR SPECIFIC APPLICATION FIELDS; PHYSICS

The invention discloses a method for predicting cytotoxicity of disinfection by-products, in particular to a method for predicting cytotoxicity of DBPs based on a machine learning algorithm by means of molecular structures and physicochemical properties of compounds. The method comprises the following steps: collecting cytotoxicity values of DBPs, and establishing a database; all the DBPs are converted into SMILES, and the SMILES is converted Calculating molecular fingerprints of all DBPs samples, and standardizing and normalizing sample data; constructing a toxicity prediction model based on multiple machine learning algorithms, and selecting an optimal model; after the SMILES expression of the to-be-detected DBPs is input, the predicted cytotoxicity value of the to-be-detected DBPs is directly output. 本发明公开了一种预测消毒副产物细胞毒性的方法，借助化合物分子结构和理化性质，利用基于机器学习算法预测DBPs细胞毒性的方法。所述方法流程包括：收集DBPs的细胞毒性值，建立数据库；将所有DBPs转化为SMILES；计算所有DBPs样本的分子指纹，对样本数据进行标准化、归一化；基于多种机器学习算法构建毒性预测模型，选出最优模型；输入待测DBPs的SMILES表达式后，直接输出待测DBPs的预测细胞毒性数值。