Tsingke Launches Workflow for High-Throughput Validation of AI-Designed Proteins

BEIJING, Sept. 22, 2026 /CourierPR/ -- Tsingke, a biotechnology company based in Beijing, has introduced a comprehensive workflow designed to accelerate the validation of AI-designed proteins and antibodies. The new system aims to streamline the experimental evaluation of large candidate libraries generated through advanced computational design tools, such as AlphaFold and ProteinMPNN.

According to Nan Zhang, the Marketing Manager at Tsingke, the workflow is crucial for researchers who are rapidly expanding their candidate pools.

The workflow consists of several stages, starting with high-throughput candidate generation and screening. Tsingke uses batch gene synthesis and parallel protein or antibody expression to move large candidate sets into experimental testing. Initial screening, including ELISA-based binding assays, helps identify candidates with the desired expression or binding characteristics.

For the second stage, Tsingke focuses on prioritized candidates. Selected molecules undergo scale-up expression, purification, and gram-scale production. These molecules are then subjected to quantitative binding analysis using technologies such as Biacore (BLI) or Surface Plasmon Resonance (SPR). These assays provide key kinetic parameters, including dissociation constant (KD), association rate (kon), and dissociation rate (koff), which support more detailed candidate characterization.

Tsingke's capabilities span gene synthesis, protein and antibody expression, and downstream characterization. The company can handle a broad range of sequence complexities and scales, including DNA fragments up to 200 kb and challenging constructs with high GC content, tandem repeats, palindromic regions, and other complex features. The GeneOptimizer algorithm optimizes codon usage to improve compatibility between gene design and downstream expression systems.

Tsingke also offers multiple expression systems, including mammalian cell-based and cell-free platforms, to match different protein characteristics, expression challenges, and project stages. This flexibility allows researchers to select an appropriate expression strategy for diverse AI-designed proteins. For antibody expression, Tsingke works with multiple formats, including scFv, VHH, Fab, and full-length IgG. Its high-throughput workflow can process up to 1,500 candidates per day, with gene-to-antibody delivery in as little as 7 calendar days, plus 5 days for shipping, for applicable projects.

Tsingke has already supported experimental projects from clients working in AI-designed proteins and antibodies. The company plans to continue enhancing its technical and service capabilities, aiming to accelerate the transition of AI-designed molecules from computational design to experimental validation.

This workflow addresses the growing need for efficient experimental validation in the rapidly advancing field of AI-driven protein design. By enabling researchers to screen candidates at scale and focus resources on the molecules with greater potential, Tsingke's system is poised to significantly streamline the development process.