Jump to content

An Automated High Throughput Engineering Platform for AI-Supported Developability Predictions

Bio-IT World, Boston
April 15, 2024

Combinatorial selection strategies and advances in protein and nucleotide engineering have been successful in generating novel large-molecule therapeutics. Bi- and multi-specific antibodies, antibody drug conjugates (ADCs), chimeric antigen receptors (CARs), engineered T-cell receptors (TCRs), and other formats offer new approaches to treatment. However, the efficient design, production, and multi-dimensional characterization represent a major challenge, especially when creating those highly engineered therapeutic candidates in high throughput. Here, we demonstrate how the Genedata Biologics® platform enables a fully automated workflow for next-gen modalities, integrating all steps from selection, molecular biology, expression, purification, and analytics. Built-in workflows for automated in silico molecule assembly mechanisms allow efficient design of large panels of novel biomolecules. Dedicated tools for pooled cloning deconvolution and automated chain pairing recovery automate the generation of tens of thousands of molecule variants that are then tested for drug-like properties. Data from multi-parametric screening is captured in the system’s highly structured database and systematically analyzed to evaluate the candidates under consideration of all meta-data, genomic, and phenotype information. We demonstrate the platform’s capabilities by illustrating a fully integrated developability and manufacturability assessment using a novel AI/ML approach for large panels of bi-specific molecules. Further, we illustrate how the platform can be used to automate the full range of innovative modalities, including CAR-Ts, AAVs, and mRNA-LNPs.

Request Resource

By submitting my data, I give consent to the collection, processing and use of my personal data in accordance with the Genedata privacy policy