Computational tools play a pivotal role in streamlining design-to-structure workflows, and at PUXANO, we are driving innovation in this area through the development of specialized software. Our tools integrate seamlessly with open-source platforms like AlphaFold 2, RF Diffusion, ProteinMPNN, and Relion, and are tailored specifically for protein design and structure determination. One such tool is procondor, a software solution that enables the retrieval, analysis, and integration of data from various public and proprietary databases, providing comprehensive insights into specific protein targets. These insights inform multiple protein design strategies, whether focused on preserving the native sequence or optimizing functional regions, all pursued in parallel. Each strategy undergoes rigorous in silico selection, significantly minimizing the need for extensive experimental validation in our Biochemical Services.

Additionally, we are at the forefront of AI-driven tools for de novo protein structure prediction therebyintegrating experimental data, or diffusion-based protein design methods with as focus creating site-specific antigens or soluble analogues of membrane proteins. These tools are often developed in collaboration with academic and industrial partners under the EuroHPC and OpenFold open-source frameworks, as we believe sustainable and competitive growth in AI protein prediction requires collective open-source development of large-scale models and that the differentiation lays in the data and specific application optimization.

Lastly, we are contributing to the development of automated image processing software for single-particle and nanoparticle analysis, aiming to establish a fully automated pipeline for processing cryo-EM data. This initiative enhances our throughput for resolving protein structures and performing quantitative analysis, aligning seamlessly with our innovations in cryo-electron microscopy grid development.