This repository contains data and scripts accompanying the publication Benchmarking AI-Driven Protocols for Cyclic Peptide Conformer Prediction and Ranking.
The Dataset directory contains reference files - publicly available cyclic peptide structures - reproduced for the reader's convenience, along with a .csv file indexing them.
The Results directory contains structure predictions for all systems included in the Dataset, for all tools and parameter settings tested in the article. It also includes a .csv file with calculated properties corresponding to the structures (RMSD, backbone dihedrals, dRMSD, model confidence metrics) which serves as the input to the plotting scripts included in the repository.
The included Python scripts can be used to generate the figures included in the article and its Supporting Information from the provided results.csv file.
