AI-assisted and structure-based small-molecule aptamer discovery: From SELEX to validation
TL;DR
A proposed validation-aware framework aligns claims with the endpoint examined and prevents computational scores, sensor responses, or concentration traces from being interpreted as proof of affinity, mechanism, or clinical utility.
Abstract
Small-molecule aptamer discovery combines systematic evolution of ligands by exponential enrichment (SELEX), high-throughput-SELEX, artificial intelligence (AI)-assisted prioritization, structural modeling, experimental binding assays, and biosensor engineering, but these stages generate non-interchangeable evidence. This review examines how target presentation, analog chemistry, sequence analysis, AI, docking, molecular dynamics, and assay architecture support different claims. Representative small-molecule and therapeutic-monitoring studies illustrate routes from candidate generation to direct binding, selectivity, biosensor evaluation, matrix testing, and continuous monitoring. The proposed validation-aware framework aligns claims with the endpoint examined and prevents computational scores, sensor responses, or concentration traces from being interpreted as proof of affinity, mechanism, or clinical utility.