deepCDR's patented mammalian display platform supports the screening of natively expressed antibodies for any development parameter
Knowing what is present in either an immune repertoire or variant library is half the battle. We can characterize your repertoires computationally to inform decisions for moving forward
The complementary power of mammalian display and deep learning enable deepCDR to locate the optimal sequence space for your target antibodies
Traditional antibody optimization makes incremental sequence mutations from a lead candidate to find improvements for a given parameter. This approach risks finding only the local maxima, and misses out on a vast landscape of variants that will never be reached as they lie on the other side of low parameter score valleys.
deepCDR's unique cutting-edge approach weds the strengths of mammalian-display and machine learning to scan the breadth of the entire fitness landscape, defining not only the peaks and valleys for your set of parameters but locating the sequence spaces and variants with the most potential.
Mammalian display allows for the functional and behavioural screening of antibody variant libraries in a native expression environment. This results in the shortest development requirements for converting the variable regions into full length antibodies, because it's already included in the screening process, something that is not possible with other display technologies.
deepCDR therefore utilizes a proprietary mammalian display platform providing rapid-turnaround from screening to expression of interesting variants. Our technology is built around speed and efficiency - it allows us to rapidly characterize the antibodies as both membrane-bound BCRs and secreted Ig's.
Computational Repertoire Screening
Knowing which variants you have in your pool or library, and linking data concerning those variants is at the root of deepCDR's advantage. Searching the full breadth of the sequence landscape means the resulting antibody variants can be focussed around absolute maxima instead of local maxima. The resulting antibodies have the best potential to reach peak performance parameters.
Machine Learning Driven Optimization
Machine learning guides our discovery and optimization process, leading to variants that carry the highest potential for success with any set of functional or developability parameters.
deepCDR's complementary mammalian display and machine learning technologies ensure the diversity present in our screening campaigns is far beyond that normally found in mammalian display technologies, and even rivals yeast and phage display. Screening the entire sequence space enables us to locate the regions with absolute optimal fitness.