Ai-guided variant prioritization with plant-based parallel protein expression platform

VELOZBIO offers an AI-guided protein variant prioritization platform coupled with a plant-based parallel expression system, designed to accelerate the discovery and development of high-value recombinant proteins. By leveraging the natural machinery of germinating yellow pea tissue, this technology provides a more biologically relevant expression environment compared to traditional microbial or cell-free systems, particularly for proteins that require complex folding or post-translational modifications. The platform integrates computational prediction with high-throughput parallel expression to efficiently screen large variant libraries, reducing the time and cost associated with conventional protein engineering workflows.

Overview

VELOZBIO's platform addresses a critical bottleneck in protein engineering: the disconnect between computational design and functional validation in a relevant biological context. Our solution combines two core capabilities: an AI-driven prioritization engine that identifies the most promising candidates from large mutation spaces, and a parallel expression system that rapidly produces and tests these variants in germinating yellow pea tissue. This integrated workflow ensures that high-performing protein variants are not only designed in silico but are also validated in a plant-based expression environment that mimics the conditions required for structural integrity and function. The result is a streamlined path from sequence design to functional protein, enabling partners to explore broader sequence space, reduce the burden of downstream purification and screening, and accelerate the development of novel proteins for applications in food, nutrition, cosmetics, and specialty chemicals.

Technical specifications

Key capabilities:

• AI-driven variant prioritization: Our machine learning platform analyzes sequence-function relationships to rank variants by predicted performance, focusing wet-lab efforts on the most promising candidates.
• Parallel plant-based expression: Our germination platform enables the simultaneous expression of multiple protein variants (10-50 per run) in individual batches of germinating yellow pea tissue, facilitating high-throughput screening without the need for bioreactors.
• Biologically relevant expression environment: Expression in germinating plant tissue preserves the protein folding pathways and post-translational modification machinery of plant cells, which is crucial for the production of complex, functional proteins.
• Integrated workflow: Our platform provides a seamless pipeline from in-silico design and prioritization to parallel expression, crude extraction, and delivery of material for downstream activity assays.
• Rapid iteration: The parallel nature of our expression system, combined with our AI-driven design cycle, enables multiple rounds of build-test-learn within a single project timeline.
• Proprietary germination expertise: Our core technology, based on controlled germination processes, activates natural enzymes that inherently alter protein configurations, enhancing solubility and functionality in our expression system.

Technology readiness level

This technology is currently at a validated prototype stage. VELOZBIO has successfully demonstrated the platform's core capabilities through a case study involving the expression of human-identical tropoelastin, a structurally complex recombinant protein. This was achieved in 14 weeks, confirming the platform's speed and capacity for handling proteins with intricate folding requirements. The next phase involves formal technology validation with partners through a structured pilot program. This program will focus on defining target protein classes, adapting the platform for multi-variant parallel expression of partner-specific sequences, and integrating our AI prioritization pipeline with partner data systems. A successful pilot is expected to be completed within 12 months, leading to the establishment of a commercial access model for the adapted expression workflow.