Examination of a blood-derived peptide library for HIV-1 inhibitors identified a small fragment of human serum albumin, named EPI-X4, as a potent and specific inhibitor of CXCR4 signaling. This project will clarify whether EPI-X4 represents the long-sought gatekeeper preventing sexual transmission of CXCR4-tropic HIV-1 strains. CXCR4 is a promising drug target for tumor therapy and optimized derivatives of EPI-X4 show promising therapeutic effects in mouse models. The project will combine machine learning applications, biomolecular modeling, and functional assays to further optimize EPI-X4 derivatives and to evaluate their therapeutic efficacy against different types of cancers and inflammatory diseases.