Advances in nanotechnology will be necessary to help deliver editing cargo beyond easy-to-access tissue (such as blood cells) and to devise methods for tracking and controlling edited cells in vivo. Lipid, gold and polymer-based nanoparticles are now in development. Nonviral delivery is being tested in vivo in animals. Synthetic circuits are being engineered to turn off editors inside the cell if they are going off-track. This strategy could drastically limit immune reaction. Nascent efforts are underway to exert direct electrical control over the bioelectric signalling methods upstream of gene expression; pre-clinical studies show this method can control glycemic levels in diabetic mice.¹¹

There is also a need for automated analysis of human tissue. Limitations in current understanding of the complex interactions between genetic and epigenetic factors that drive many disease pathologies could be overcome by artificial intelligence — specifically machine learning algorithms — that can identify the relationships among genes, gene networks and other factors involved in disease, and the potential consequences of edits to these.¹² Machine learning may also be able to help identify novel biological candidate systems to manipulate DNA: it would be useful to find molecules that offer decreased immunogenicity, for instance. Searching through microbial data obtained from uncultivated samples may reveal more suitable enzymes — helicases, nucleases, transposases or recombinases — that solve the problems of currently available editors.