Decoy Therapeutics, Inc. announced a collaboration with Quantori, a provider of digital transformation services and technology for the life sciences and healthcare industries, supported by funding from Google Cloud, to build a Google Cloud-native peptide design platform intended to dramatically increase the speed and scale of antiviral peptide discovery. The collaboration will deliver a production-grade, cloud-based infrastructure built on Google Cloud Platform that integrates advanced protein folding models, molecular dynamics simulations and performance optimization tools into a single scalable environment. The new platform is designed to integrate Cloud-native infrastructure, advanced machine learning and high-performance computing directly into Decoy's proprietary IMP³ACT design-build-test-learn engine.
By unifying these capabilities into a single production-grade computational environment, Decoy expects to increase computational throughput by approximately 25- to 100-fold, enabling a step-change improvement in how rapidly new peptide candidates can be conceived, optimized and advanced. Additionally, the modularity of the platform will allow Decoy to continually adapt and upgrade its capabilities in the fluid competitive space of AI-driven molecular design. The architecture also lays the foundation for a new generation of AI-enabled research tools that can integrate emerging protein design models, generative AI techniques, and experimental data streams into Decoy's discovery workflows.
Under the agreement, Quantori will deploy a fully containerized, GPU-accelerated computational architecture powered by components of its Q-Suite engineering platform. These accelerators enable rapid integration of AI models, scalable high-performance computing workflows, and production-grade scientific applications tailored to the needs of computational drug discovery. The platform will support tiered protein folding pipelines, enabling dynamic trade-offs between speed and structural accuracy, advanced molecular dynamics simulations with GPU acceleration and AI-enhanced sampling, diffusion-based conformational ensemble generation to improve structural plausibility, automated algorithm selection and precision optimization to balance computational efficiency and model fidelity, and real-time monitoring, visualization of 3D structures and performance benchmarking through a unified interface.
The unified environment will support forward and reverse protein folding workflows, enabling both structure prediction and design applications. Integration with molecular dynamics pipelines will allow Decoy to move seamlessly from sequence generation to structural validation and conformational analysis within a single orchestrated system. The initiative builds on Decoy's existing momentum in applying an engineering mindset to drug discovery and reflects the Company's broader strategy of leveraging advanced computation to reduce the time, cost and risk associated with developing peptide-conjugate therapeutics.
