Diffuse Bio, a leading provider of AI and experimental platforms for designing and screening protein therapeutics, announced significant enhancements to DiffuseSandbox, its self-serve protein design platform. DiffuseSandbox now offers large-scale binder design capabilities and direct integration with RamaX, the company's rapid binder screening platform.

The updates create a comprehensive, end-to-end system that enables protein binder discovery in 1-2 weeks—where users can design and then immediately submit large libraries of AI-designed binders such as minibinders, nanobodies, and scFvs for wet lab validation, all within a few clicks.

From Binder Design to Validation in One Step

By directly connecting DiffuseSandbox, its AI design platform, to RamaX experimental screening, Diffuse Bio has created a seamless workflow to design and screen large libraries of protein binders against up to 100 different targets simultaneously—all within 1-2 weeks.

This combination of speed, scale, sensitivity, and selectivity transforms what was once a months-long, resource-intensive, and failure-prone process into a streamlined operation that can be initiated with just a few clicks—enabling researchers to quickly move from initial design concept to validated protein binders in just one step.

Key Platform Enhancements

The expanded DiffuseSandbox now includes two major capabilities:

• Large-scale design generation: Request up to 10,000 VHH, scFv, or minibinder designs in a single run, enabling comprehensive library generation with complete molecule ownership.
• Integrated RamaX screening: Upon job completion, users can request quotes to screen up to 100,000 designs against up to 100 antigens in parallel using RamaX, streamlining the path from computational design to experimental validation.

"RamaX revolutionizes protein therapeutic and reagent design by enabling the rapid discovery of protein binders for up to 100 antigens in parallel, compressing months of work into 1-2 weeks,” notes Namrata Anand, CEO at Diffuse Bio. “This unprecedented speed and scale will accelerate drug programs and will also allow us to generate massive datasets to train better AI protein design models. By rapidly generating large, accurate datasets, we are building the critical foundation required to scale AI performance and unlock the future of generative protein design."