The de novo design of symmetric protein nanocages has emerged as a major frontier in programmable biomolecular engineering due to its potential applications in nanomedicine, vaccine development, molecular encapsulation, catalysis, and synthetic cellular systems. Despite substantial progress in computational protein design, the reliable construction of large polyhedral assemblies remains limited by kinetic mis-assembly, off-pathway oligomerization, interface instability, and low experimental success rates. This study presents a computationally integrated framework for the design of programmable icosahedral protein nanocages with improved assembly fidelity. The proposed strategy combines symmetry-guided scaffold selection, Rosetta-based symmetric docking, RFdiffusion-driven interface backbone generation, ProteinMPNN sequence optimization, and AlphaFold2-Multimer computational filtering to produce experimentally tractable cage architectures. The target system consists of a two-component assembly formed from trimeric and dimeric oligomeric building blocks organized into a 60-subunit icosahedral nanocage approximately 25 nm in diameter. The framework emphasizes reduction of non-specific intermolecular interactions through interface-specific design constraints and computational pre-screening of unstable configurations. Predicted outcomes indicate that the integration of diffusion-based backbone generation with sequence-level optimization and AF2-based structural validation substantially improves the likelihood of obtaining stable self-assembling cages while reducing experimental screening burdens. The study further evaluates how this integrated workflow compares with existing approaches in symmetric protein cage engineering. Collectively, the proposed methodology provides a scalable route toward high-fidelity programmable protein nanomaterials and contributes to the broader development of de novo self-assembling biomolecular systems.

Solid-state batteries promise enhanced safety, higher energy density, and improved thermal stability by replacing flammable liquid electrolytes with solid-state electrolytes (SSEs). This study presents a systematic framework for ranking SSE candidates for lithium batteries. After surveying fundamentals of solid-state batteries and common SSE classes, a comprehensive set of performance characteristics was derived from literature. These characteristics were used as a basis for a concise set of parameters specific to the Materials Project (MP) database. High-throughput screening of Li-containing compounds (binary to octonary) in the MP database generated a large pool of candidates. These were compiled into a dataset and ranked via a hybrid supervised-unsupervised machine-learning (ML) workflow calibrated on literature benchmarks. The resulting ranked list shows good overall qualitative agreement with experimental data on key SSE candidates, but with some notable discrepancies. The top-ranked candidates include garnet oxides, anti-perovskites, sulfides, halides, and NASICON-type materials, each exhibiting distinct trade-offs between conductivity, stability, and processability. The ranking aligns closely with prior high-throughput and ML-assisted screening studies while providing improved multi-objective integration. Major strengths of the approach in this study include its reproducibility and comprehensive property balancing. The transparent methodology of this study accelerates discovery and optimization of SSEs, advancing safer and higher-energy-density lithium solid-state batteries for electric vehicles, grid storage, and portable applications.