Protein-nucleic acid (NA) interactions are central to numerous biological processes and to biotechnology. Yet, it is still difficult to predict structures and specificities of natural protein-NA complexes in silico, limiting understanding of their structure-function relationships and the ability to bioprospect valuable nucleic acid-binding proteins. When it comes to design tasks, scientists have been restricted to the repurposing of natural NA binding proteins through domain fusions, small sets of rational or laboratory-evolved mutations, and conservative manipulation of structures. This limits the potential application space and the efficacy of solutions; as examples, there are no programmable solutions for binding disease-relevant higher-order DNA or RNA structures and writing multi-kilobase segments of DNA into genomes remains challenging. Our lab’s goal is to develop new capabilities for computational prediction and design of protein-NA assemblies and apply these capabilities to address real-world problems. To do this we use a variety of tools including both physics-based and AI/ML approaches for protein design/modeling, high-throughput biochemistry, molecular biology, and sequencing technologies. Of particular interest is the development of cutting-edge protein functional assays to augment training of AI design models in low-data scenarios.