Today the clinical application of many otherwise promising new strategies for liver directed gene and cell therapy is hampered by their inherent quantitative inefficiency. Here we propose to use in vivo selection to overcome these problems. We will develop clinically applicable ways to use small molecule drugs to select genetically modified hepatocytes in vivo and thereby cross the quantitative threshold required for permanent therapeutic benefit. We have identified the fever medication acetaminophen (APAP) as excellent candidate for this approach. Three specific aims are designed to fully validate these novel method for in vivo selection of therapeutically modified hepatocytes. In Aim 1, we will develop integrating rAAV vectors (Generide vectors) that can be selected by APAP and in aim 2,conditionally selectable lentiviral vectors will be generated. Both will be validated in metabolic disease models. In Aim 3, we will apply the results of Aims 1&2 to human hepatocytes. Selection of genetically modified human hepatocytes will be carried out in liver chimeric mice.