A Disorder-Aware Computational Framework to Identify Structurally Tractable Targets in Proliferative...

Area of Science:

• Ophthalmology
• Computational Biology
• Drug Discovery

Background:

• Proliferative vitreoretinopathy (PVR) is a leading cause of vision loss after retinal detachment surgery.
• Current treatments for PVR lack efficacy, and pharmacologic options are limited.
• Epithelial-mesenchymal transition (EMT) is a key driver of PVR, but its regulators are poorly understood.

Purpose of the Study:

• To develop a computational framework to identify and prioritize drug targets for PVR.
• To evaluate EMT-associated proteins for structural tractability using artificial intelligence (AI).
• To enable structure-based drug design for PVR treatment.

Main Methods:

• A computational pipeline was used to screen 25 EMT-associated proteins implicated in PVR.
• Methods included intrinsic disorder profiling, redox-sensitive disorder-to-order transition (DOT) analysis, and protein-protein interaction network assessment.
• Structure-based modeling and AI-driven binder design (RFdiffusion, AlphaFold2) were applied to prioritize candidates.

Main Results:

• The framework identified several proteins with intrinsic disorder and DOT potential.
• Snail Family Transcriptional Repressor 1 (SNAIL1) was prioritized due to its disorder profile, DOT potential, and network connectivity.
• A de novo binder targeting SNAIL1's disordered region was designed with high structural confidence.

Conclusions:

• A novel AI-driven computational framework can prioritize structurally tractable EMT regulators for PVR.
• SNAIL1 is a promising therapeutic target for PVR.
• The study provides a foundation for developing disorder-based therapies for fibrotic retinal diseases.