Schematic of the computational approach for identifying an HS-CRM8. (A) Overview of steps involved in HS-CRM8 identification. Human liver tissue was used to identify genes that are specifically highly or lowly expressed by using microarrays. Next, the respective promoters for the highly and lowly expressed genes were extracted from existing databases. Transcription factor binding sites (TFBSs) were then annotated by using the TRANSFAC database and were subjected to a computational algorithm to identify HS-CRM8, which was further refined by analysis of evolutionary conservation. (B) Schematic of the best performing scAAV9 vector used in the study with a codon-optimized human factor IX Padua mutant (cohFIX R338L) expressed from the HS-CRM8 and a minimal transthyretin (TTR) promoter.