An Optimized Modeling Algorithm for Breast Cancer Drug Candidates Based on NSGAII
DOI:
https://doi.org/10.54097/fbem.v2i3.193Keywords:
ResNet residual network, NSGAⅡ multi-objective optimization algorithm, Choquet fuzzy integral; bioactivity prediction, Neural networkAbstract
Breast cancer is one of the most common malignant tumors in women. It seriously threatens the safety of women worldwide. It is an important and urgent task to research and develop anti-breast cancer drugs and improve the therapeutic effect of breast cancer. Taking the actual sample data as the main starting point, firstly, the prediction model of pIC50 is established by ResNet residual network and neural network (NN) to judge the biological activity. Then the classification model of ADMET property is established by ResNet residual network and LightGBM, and the model fusion is realized by Choquet fuzzy integral. Finally, the NSGAII multi-objective optimization algorithm is used to determine the range of values that each molecular descriptor obtains in the range of good biological activity, and ultimately to optimize the modeling of anti-breast cancer drug candidates. The experimental results show that the algorithm improves the prediction accuracy of biological activity, realizes the efficient and accurate classification of ADMET properties, and accurately describes the impact of molecular descriptors on biological activity.
Downloads
References
Ni Wenting. Exploring the effect of cryptotanshinone on ERα+ based on BCRP Molecular mechanism of proliferation inhibition in breast cancer [D]. Nanjing University of Chinese Medicine, 2019.
Deb K, Pratap A, Agarwal S, et al. A fast and elitist multiobjective genetic algorithm: NSGA-II[J]. IEEE transactions on evolutionary computation, 2002, 6(2): 182-197. DOI: https://doi.org/10.1109/4235.996017
He K, Zhang X , Ren S , et al. Residual for Image Recognition[J]. IEEE, 2016. DOI: https://doi.org/10.1109/CVPR.2016.90
Meng Q. LightGBM: A Highly Efficient Gradient Boosting Decision Tree. 2018.
Islam M A, Anderson DT, Pinar A J , et al. Enabling Explainable Fusion in Deep Learning with Fuzzy Integral Neural Networks[J]. IEEE Transactions on Fuzzy Systems, 2019. DOI: https://doi.org/10.1109/TFUZZ.2019.2917124
