Artificial Intelligence-Mediated Computer-Aided Design of Viral Gene Therapies

Abstract

Over 5% of newborns suffer from a genetic disease. These include single gene, polygenic, and chromosomal disorders. Many other noncongenital diseases with genetic components are activated by environmental triggers (autoimmune, cancer, and tissue injury). Sophisticated viral gene therapies could treat, and possibly cure, these diseases and significantly ease patient burden and improve quality of life. Current viral therapies are mostly limited to plasmid-based and adeno-associated virus variants with inefficient response rates and limited use, with some herpes, lenti, and retroviral modalities. Development is slow and expensive. Virtual prototyping of viral gene therapies through computational design, like in other engineering fields, may represent a useful process to accelerate and expand viral pipeline development by opening the human virome to therapeutic development and constructing specificity, potency, efficacy, and safety in silico. Contemporary computational tools (artificial intelligence, machine and deep learning, computer-aided design, high performance computing, cloud and edge computing, and physics-based modeling) now render this possibility feasible and, therefore, constitute powerful options for biopharma researchers to expand and accelerate precision medicine research and development for complex indications.