Background: Age-related macular degeneration (AMD) is a significant factor causing blindness in adults. However, the clinical diagnosis of AMD is relatively challenging, due to the shortcomings of the existing clinical examination methods and the latent period of retinal damage before macular degeneration becomes apparent. This study aims to explore the potential of extracellular vesicles (EVs) protein chips for early diagnosis of AMD using patients' plasma samples.
Methods: To achieve early diagnosis of AMD, this study utilized a high-throughput platform for liquid biopsy based on EVs protein chips. Forty AMD patients and 41 normal individuals were recruited. Through machine learning methods, we identified that ATP-binding cassette transporter A1 (ABCA1) is an EVs protein marker for diagnosing AMD. Additionally, a validation set was constructed using the random forest method for verification.
Results: The results of the study indicated that ABCA1 is a reliable biomarker for diagnosing AMD. The validation using the random forest method confirmed the robustness and reliability of ABCA1 as a diagnostic marker. This finding suggested that ABCA1 can serve as a new promising liquid biopsy-based marker for diagnosing macular degeneration.
Conclusion: The utilization of EVs protein chips, combined with machine learning methods, can effectively identify ABCA1 as a biomarker for the early diagnosis of AMD. This approach offers a promising new method for liquid biopsy diagnostics, potentially improving the clinical diagnosis and management of macular degeneration.
Keywords: age-related macular degeneration; extracellular vesicles; liquid biopsy diagnostics; machine learning.