Ligand Pharmaceuticals and SandboxAQ: Revolutionizing Drug Discovery with SAIR Dataset

### Transforming Drug Discovery Through AI: The Launch of SAIR

On June 18, 2025, SandboxAQ, a leading AI-driven company based in Palo Alto, California, unveils the SAIR (Structurally Augmented IC50 Repository), a groundbreaking dataset that is set to revolutionize the field of computational drug discovery. This repository stands as the largest collection of protein-ligand pairs with detailed experimental potency data to date. The introduction of SAIR is expected to significantly enhance the speed and precision of binding affinity predictions, providing researchers with a crucial tool to further develop AI models in drug discovery. As the industry shifts towards more data-centric approaches, the implications of this launch extend beyond traditional methodologies, promising to streamline the drug development process.

The creation of over 5.2 million synthetic 3D molecular structures across more than one million protein-ligand systems showcases SandboxAQ's commitment to innovation. Utilizing NVIDIA DGX™ Cloud for AI training, the company achieves a twofold improvement in GPU utilization, which enhances throughput for scientific workloads. This collaboration not only demonstrates the power of advanced computational resources but also highlights the synergy between cutting-edge technology and scientific exploration. Nadia Harhen, General Manager of AI Simulation at SandboxAQ, describes the SAIR dataset as a transformative milestone in drug discovery, enabling a shift from conventional trial-and-error methods to a more rapid, data-driven approach.

SAIR's comprehensive dataset allows researchers to predict protein-ligand binding affinities at speeds that are at least 1,000 times faster than traditional methods, significantly expediting the journey from drug discovery to market readiness. By making this wealth of data publicly available, SandboxAQ aims to empower scientists to train advanced models quickly and efficiently, thereby altering the landscape of drug discovery. The accessibility of SAIR marks a pivotal advancement in the scientific community, enabling researchers to leverage AI in a manner that was previously unattainable. As the industry embraces these innovations, the potential for accelerated drug development grows, promising new treatments and therapies on the horizon.

In related developments, the release of the SAIR dataset is accompanied by a preprint available on bioRxiv, titled "SAIR (Structurally Augmented IC50 Repository): Enabling Deep Learning for Protein-Ligand Interactions with a Synthetic Structural Dataset." This resource serves as an additional reference for researchers eager to understand the implications of the dataset and its applications. SandboxAQ's innovative quantitative AI technology has already demonstrated its ability to deliver superior results in the field of drug discovery, setting a new standard for future research endeavors.