Novel Scoring Methods for Interactions between Antibodies and Antigens

我们将结合联合收割机最新的自由能计算方法和增强的采样分子动力学模拟来预测抗体与靶蛋白表位结合时的构象，并随后通过亲和力对不同的环序列进行排序。在UCB，我们已经从历史计算抗体设计中积累了大量的实验数据，用于概念验证靶蛋白。这些数据将用于验证项目期间开发的评分方法，以检测已知结合和非结合抗体设计的区分。此反馈将用于迭代改进评分方法

We will combine the latest free energy calculation methods and enhanced sampling molecular dynamics simulations to predict antibody conformations when bound to the target protein epitope, and to subsequently rank the different loop sequences by affinity. At UCB, we have accumulated a large body of experimental data from historical computational antibody designs, for a proof-of-concept target protein. These data will be used to validate the scoring method developed during the project to test for discrimination of known binding and non-binding antibody designs. This feedback will be used to iteratively improve the scoring method