Advancing ocular gene therapy: a machine learning approach to enhance delivery, uptake and gene expression

Abstract

Ocular gene therapy offers a promising approach for treating various eye diseases, centered on the process of transfection, including delivery, cellular uptake and gene expression. This study addresses anatomical and physiological barriers, such as the eyelids, tear film, conjunctiva, cornea, sclera, choroid and retina, affecting therapeutic success. A three-step machine-learning approach is proposed. The first step predicts gene delivery efficacy by integrating molecular characteristics of the ocular gene therapy product, ocular barrier properties and patient demographics. The second step predicts cellular uptake rates, analyzing product penetration and cellular interactions. The final step forecasts gene expression levels, considering factors like nucleic acid type and endosomal escape. An artificial neural network model is recommended to capture complex, nonlinear relationships, enhancing our understanding of therapeutic and biological interactions.

Keywords: ANN; ML; Ocular gene therapy; artificial neural networks; cellular uptake; gene delivery; gene expression; machine learning; ophthalmology.