Docking of 8400 peptides into an experimental structure of bacterial peptide transporter DtpB

Di- and tripeptides composed of proteinogenic amino acids were docked into an experimental structure of DtpB using Rosetta FlexPepDock protocol. Additional experimental structures of DtpB were used to train a classifier to select top docking poses. Average Rosetta energy terms for top 10 hits were then combined with experimental affinity measurements to train a binary classifier that predicts binders to DtpB.

This upload includes input and output data as well as code to run and analyse the full pipeline. Some intermediate results are available upon request.

For more details, please see:
Plasticity of the binding pocket in peptide transporters underpins promiscuous substrate recognition

https://dx.doi.org/10.1016/j.celrep.2023.112831

采用Rosetta FlexPepDock协议，将由蛋白源性氨基酸构成的二肽与三肽对接至DtpB的实验结构中。利用额外获取的DtpB实验结构训练分类器，以筛选最优对接构象。随后将Top10命中结果的平均Rosetta能量项与实验亲和性测定数据相结合，用于训练可预测DtpB结合物的二分类器。

本上传数据集包含完整流程的输入与输出数据，以及用于运行和分析全流程的代码。部分中间结果可按需提供。

更多细节请参阅：《肽转运蛋白结合口袋的可塑性支撑混杂底物识别》
https://dx.doi.org/10.1016/j.celrep.2023.112831