Within EYEGET we were able to significantly improve the ability of delivering large genes via AAV vectors to the retina. We have set-up a strategy that uses 3 AAV vectors at the same time which upon recombination in target cells can reconstitute therapeutic DNA which is about 3 times larger than what normally carried by a single AAV. In addition, as this system based on AAV DNA recombination is not efficient enough for some applications, we have designed a new system based on protein trans-splicing, in which two polypeptides flanked by split-inteins fuse together through the inteins which then excise seamlessly from the full length mature protein (a cut-and-paste at the protein level). We have shown that AAV-intein reconstitute large therapeutic proteins defective in IRDs like Stargardt disease or Leber congenital amaurosis 10 to levels that are therapeutic in mouse models of these diseases. We have also implemented AAV-based genome editing to specifically target common Rhodopsin mutations which cause dominant retinitis pigmentosa, and we have demonstrated the efficacy of this strategy in mouse models. In addition we have set-up a novel mutation-independent genome editing strategy, i.e. than can be applied to any Rhodopsin mutation, that is based on homology-independent targeted integration of a donor DNA at the Rhodopsin locus. This approach, based on subretinal delivery of AAV, results in therapeutic efficacy in a mouse model of dominant retinitis pigmentosa. Finally, we have developed human 3D retinas in a dish which contain cells which resemble photoreceptors and we have successfully tested some of the gene therapy platforms developed within EYEGET in this human model.