Lung cancer remains a leading cause of cancer-related mortality worldwide, highlighting the urgent need for rapid, accurate, and affordable diagnostic strategies. UBA6-specific E2 conjugating enzyme 1 (USE1) is overexpressed in lung cancer and contributes to tumorigenesis, yet no clinically applicable method exists for its detection. We developed an AI-assisted aptamer biosensing platform for antibody-free detection of USE1. High-affinity candidates (Aptamer 1e) were identified through systematic SELEX and rational truncation, and AlphaFold3-based modeling was subsequently applied post hoc to provide a structural hypothesis for the observed binding. For signal amplification and visualization, we engineered a nanostructured detection system composed of rolling-circle–amplified DNA microspheres (DNAMS) conjugated with streptavidin–quantum dots (STA-QDs). The DNAMS–STA-QD biosensor enabled strong fluorescence signals in USE1-positive cancer cells and produced a clear visual distinction between tumor and matched normal lung tissues. In 30 paired tissue samples, the biosensor achieved AUC = 0.961, with 86.7% sensitivity and 93.3% specificity for detecting lung cancer. The assay requires no antibodies, enzymatic amplification, or specialized instrumentation, and offers a rapid, low-cost workflow. This study presents a clinically oriented nanobiosensing platform that integrates AI-assisted aptamer structural modeling with quantum dot–enhanced DNA nanostructures for sensitive detection of USE1. The approach offers a robust, antibody-free method for lung cancer diagnosis and demonstrates the potential of combining deep learning with nanobiotechnology to accelerate biomarker detection tool development.
Kim et al. (Tue,) studied this question.