Physicochemical and RNA structural features of selective RNA binding by drug-like small molecules

Binding selectivity is critical for small molecules targeting RNA in the therapeutic field, requiring rational design of selective compounds. However, the key features of selective binding remain incompletely understood due to lack of comprehensive datasets across diverse RNA targets. Here, we systematically analyzed 295 million interactions between 79,289 drug-like compounds and 3,725 structured RNA motifs, ranging from 50 to 150 nucleotides. Chemical space visualization reveals unexplored regions where compounds achieve both RNA binding and high drug-likeness. Selective binders exhibit distinct properties including higher molecular weight, increased lipophilicity, and fewer hydrogen bond donors. Structural analysis of target RNA shows that multiloop motifs are enriched among selective binders, providing guidance for target selection. These findings establish design principles for the next generation of RNA therapeutics by defining key features enabling selective recognition within pharmaceutically relevant chemical space.