Computational Framework Proposal for Competitive Extracorporeal Binding of Protein-Bound Uremic Toxins

Protein-bound uremic toxins (PBUTs), particularly indoxyl sulfate (IS) and p-cresyl sulfate (pCS), represent a critical limitation of conventional hemodialysis (Vanholder et al., 2014; Niwa, 2010). Although current dialysis membranes effectively remove low-molecular-weight solutes (95% of the toxin circulating in a bound state, with albumin acting as an intravascular reservoir during the dialysis session. This restricts the net reduction of IS to approx. 30–40% per session, contributing to chronic inflammation, cardiovascular fibrosis, and increased mortality. This project outlines a conceptual computational strategy for the de novo design of a synthetic mini-protein (designated TRAPHOLE-UT) utilizing structural artificial intelligence tools (RFdiffusion, ProteinMPNN), pending quantitative and experimental validation. TRAPHOLE-UT will be: 1. Low molecular weight (