MOF-LENS: Bio-inspired lotus effect optimization for accelerated discovery of metal–organic framework nanocarriers for doxorubicin delivery in cancer therapy

Metal–organic frameworks (MOFs) are crystalline materials characterized by adjustable porosity and chemical properties, rendering them effective carriers for doxorubicin (DOX) in the treatment of hepatocellular carcinoma (HCC). The extensive design space complicates the identification of metal-organic frameworks (MOFs) that simultaneously optimize doxorubicin (DOX) loading, pH-responsive release, and biocompatibility. We introduce MOF-LENS (Latent Evolutionary Navigation System), a bio-inspired framework that integrates structural descriptors (pore size, surface area, void fraction) and SMILES-based fingerprints into a unified latent space. This framework employs the previously published Lotus Effect Algorithm (LEA) to identify high-quality MOF nanocarriers. In benchmarks comparing particle swarm optimization, genetic algorithms, random search, and a deterministic filter baseline across 30 independent runs, MOF-LENS achieves comparable or superior best fitness with significantly reduced variance, resulting in a consistent set of top candidates rather than isolated outliers. The chosen MOFs demonstrate pore-limiting diameters within the DOX-compatible range (12–16 Å) and exhibit a strong chemical affinity for DOX. Docking calculations validate substantial binding at physiological pH, with interactions slightly diminished under mildly acidic conditions, aligning with controlled release mechanisms. While illustrated for DOX in HCC, MOF-LENS is applicable to various drugs; retargeting to alternatives like paclitaxel necessitates merely the modification of the drug fingerprint, pore-size window, and a limited set of fitness weights.