CDDLeiden/COFOLDER: Version 0.1.0

Zenodo (CERN European Organization for Nuclear Research)

Abstract

COFOLDER is an open-source Python framework for evaluating and applying biomolecular co-folding models in ligand screening and design. It provides Validate, Screen, and Oracle workflows that integrate model-confidence and binding outputs with structural diagnostics, robustness assessment, and estimates of protein and ligand training-set proximity. COFOLDER is described in the manuscript “COFOLDER: Toward Bias-Aware Co-Folding for Ligand Screening and Design” by Remco L. van den Broek, Bola Khalil, José Carlos Gómez-Tamayo, Gary Tresadern, Willem Jespers, and Gerard J. P. van Westen.

Authors 1

1. Remco L. van den Broek Aachen Division of Medicinal Chemistry Leiden Academic Centre for Drug Research

   Leiden University

   Affiliation as printed

   Division of Medicinal Chemistry, Leiden Academic Centre for Drug Research, Leiden University, Einsteinweg 55, 2333 CC Leiden, The Netherlands