Drug delivery based on nanoparticles faces ongoing challenges, including complex preparation processes and limited accumulation at the target site. Here, SP-13786 (SP) is introduced as a precise small molecule inhibitor that inhibits fibroblast activation protein (FAP), serving as a versatile and effective carrier, which can be easily co-precipitated with various hydrophobic drugs to form stable nanoparticles (SCAN). Screening of 861 compounds revealed that SP has a wide range of enhanced colloidal stability and drug loading capacity. Corresponding simulations and interpretable machine learning (XML) show that the assembly of SCAN depends on the balance of aromaticity, rigidity, and nitrogen-mediated interactions, providing an interpretable framework for co-assembling nanomedicine. Biological evaluations indicate that SCAN enhances drug delivery and therapeutic effects in FAP-positive cells, thereby enhancing drug penetration across the barrier induced by fibrosis and increasing drug accumulation in fibrotic tissues. The improved bioavailability is associated with excellent therapeutic effects in patients with progressive fibrosis in various disease models. Overall, we have established SP as a multifunctional neurotherapeutic platform that combines the simplicity of the preparation process, the mechanistic insights provided by XML, and its wide applicability to pathological fibrosis and drug delivery disorders. This study was published in Advanced Materials under the title "Machine Learning-Informed Nano Co-Assembly Inhibits Fibroblast Activation Protein and Improves Drug Delivery in Fibrotic Tissue".
References: DOI: 10.1002/adma.202519805