MENU

GluePlex: Predicting CRBN-Neosubstrate Ternary Complex Structures

Research Square
Overview
GluePlex improves the prediction of novel cereblon (CRBN) ternary complexes by combining restrained protein–protein docking with template-guided Boltz-2 co-folding. In tests on novel pairs, it increased the number of acceptable poses to 23 of 39 runs with a median DockQ of 0.43, compared to 12 of 39 runs and a median DockQ of 0.01 for plain Boltz-2. This approach moves molecular-glue degrader design toward a more rational, rather than empirical, process.
In Depth

Key Findings

GluePlex significantly enhances the accuracy of predicting molecular-glue induced CRBN-neosubstrate ternary complex structures. When tested on 13 novel pairs, GluePlex achieved a median DockQ (protein-interface accuracy score) of 0.43, with 19 rank-1 successes among 23 successful runs out of 39. In contrast, plain Boltz-2 co-folding produced acceptable poses in only 12 of 39 runs, with a median DockQ of 0.01.

Technical Details

The GluePlex pipeline integrates three primary components:

  • Ligand-free, restrained protein–protein docking.
  • Two passes of template-guided Boltz-2 co-folding.
  • The use of multiple conformers.

Ablation analysis indicated that the majority of the performance gains were attributed to the docking and template transfer components. However, ligand placement accuracy remains a challenge. While the CRBN-binding warhead showed a median root-mean-square deviation (RMSD) of 0.7 Å, the distal moiety had a median RMSD of 2.8 Å, remaining inaccurate even when the protein interface was correctly predicted.

Background & Context

Molecular glue degraders function through a single induced geometry. Predicting this geometry in silico is essential for transitioning degrader design from empirical methods to rational design. To evaluate performance, the researchers assembled 52 CRBN target–compound pairs with published ternary structures, stratifying them by their exposure to the training set to specifically test performance on structurally novel neosubstrates.

Significance & Outlook

Across various configurations and ligand truncations, an acceptable pose was recovered for 12 of 13 novel families, while the final family could be recovered using plain co-folding. This indicates that while no single configuration solves all cases, GluePlex broadly improves prediction capabilities. The primary remaining limitations for the technology are model ranking and the precise placement of the distal ligand.

Source: https://doi.org/10.21203/rs.3.rs-11580478/v1

Get our weekly technology intelligence — free

Receive an infographic that lets you judge at a glance whether each field’s analysis report is worth reading.

Subscribe Free — Weekly Tech Intelligence

By subscribing, you’ll receive Troy-Technical’s weekly technology intelligence newsletter.

  • Your email and selected fields are used only to deliver the newsletter.
  • We never share your information with third parties.
  • You can unsubscribe anytime via the link in each email.

See our Privacy Policy for details.

Takes about a minute · Unsubscribe anytime

Published by Troy-Technical, an independent site run by one engineer with a career in materials development.
About the author / Contact info@troy-technical.jp
Let's share this post !

Author of this article

TOC