Integrative structural profiling and ligand optimisation across the transthyretin mutational landscape

Area of Science:

• Computational biology
• Drug discovery
• Genetics

Background:

• Transthyretin amyloidosis (ATTR) is a genetic disorder from TTR protein mutations causing aggregation.
• Approved TTR stabilizers (tafamidis, acoramidis) have unclear efficacy across diverse TTR variants.

Purpose of the Study:

• To develop and validate an in silico pipeline for assessing TTR variant-specific drug binding.
• To investigate the impact of TTR mutations on approved stabilizer efficacy.
• To explore personalized drug design strategies for ATTR.

Main Methods:

• Utilized AlphaFold3 for structure prediction and ESM2 for sequence embeddings.
• Employed DiffDock-L and AutoDock Vina for molecular docking simulations.
• Applied DiffSBDD for ligand generation and embedding-based clustering for variant classification.

Main Results:

• Significant variation in ligand binding affinities across TTR variants was observed.
• Mutations distant from the binding site (e.g., W61L, Y98F) were found to reduce binding affinity.
• Customized ligand optimization showed mutation-dependent recovery of binding affinity.
• Embedding-based clustering facilitated the identification of potentially benign mutations.

Conclusions:

• ATTR therapeutic strategies must be variant-aware due to differential drug binding.
• The developed in silico pipeline supports scalable variant classification and personalized drug design.
• This approach lays the groundwork for creating tailored TTR stabilizers for individual patients.