Description
PreDBA webserver which can effectively predict the binding affinity of protein-DNA complexes using an Machine learning method. Figure 1 shows the flowchart of PreDBA. It involves four steps. (A)Get complex datasets that meet prediction conditions. (B)Classification of complexes by the type of DNA associated with the protein. (C)Extract sequence and structural features that are valid for predictive experiments. (D)Firstly, the machine learning models such as GBRT(Gradient Boosted Regression Tree), AdaR(Adaboost Regression), BagR(Bagging Regression) are used to predict the protein_DNA binding affinity, and then the predicted results are taken as new features. Finally, XGBR(XGBoost Regression) is used to predict the binding affinity based on the new features.
Current Version
* Current Version Number: 1.1 * Release Date of Current Version: 10/31/2018
Contact Info
E-mail:leideng(at)csu.edu.cn
References [1]
