Antibody humanization

Transform your antibody candidates into effective therapeutics with our deep learning aided humanization services. Whether you need expert support or a powerful self-service software solution, we provide the flexibility and precision you need to streamline your projects.

Comparing humanization approaches

Antibody humanization is a crucial step in reducing immunogenicity while maintaining therapeutic efficacy, essential for advancing a drug to the clinic. Traditional grafting methods, where CDRs are transferred from the parental molecule to human frameworks, are laborious, time-consuming, typically low throughput, and often reduce binding affinity. Deep learning has proven capable of overcoming some of these challenges, but current models like BioPhi still tend to fall short in retaining binding affinity for most generated variants.

ENPICOM’s generative deep-learning humanization model eliminates the need for traditional grafting while showing far superior performance in retaining binding affinity when compared to models like BioPhi. Achieving higher success rates with fewer screened variants increases efficiency in both time and cost. Typically, screening just 3-5 humanized variants per parental clone is already enough to find a candidate that retains binding affinity.

The ENPICOM humanization model

The significantly improved binding retention at clinically relevant humanness levels leads to considerable time and cost savings. With better retention, fewer sequences need to be screened, and sequence generation is faster. How is this achieved?

The ENPICOM humanization model is built on our proprietary paired chain protein language model (pLM), which is part of our AIGX pLM family. In addition, we have trained a generative interface to suggest mutations that optimize sequence humanness, while allowing flexibility to protect residues expected to be key for binding. Besides, the model is designed to function agnostically with regard to the parental species and is fully flexible to account for optional additional restrictions. As a result, it can be run as a self-service software solution or operated by one of our experts as a service.

Benefits of humanization powered by ENPICOM

Case study: Achieving superior success rates compared to BioPhi

In collaboration with Genovac, a leading antibody discovery company, the ENPICOM model outperformed BioPhi in antibody humanization while maintaining high germline identity and superior binding affinity. Our model can humanize antibodies from any species and offers a selection of candidates for testing, where testing just 3-5 variants typically variants is sufficient to identify a high-affinity candidate, achieving clinically relevant humanness levels (>85% and >90% germline identity) and strong affinity retention similar to the parental clone.

Getting started

We offer two customizable options, ensuring flexibility for your antibody humanization needs.

Humanization as a service

Take the complexity out of antibody humanization with our dedicated team of experts. We handle every step of the process, from optimizing sequences to ensuring they are primed for further testing and development.

Self-service humanization tool

Empower your team to humanize antibodies independently with our easy-to-use humanization tool, controlling every step of the process yourself.

Trusted by industry leaders

Revolutionize antibody humanization with ENPICOM

Schedule an exploratory meeting to explore how ENPICOM can transform your antibody humanization process and accelerate your therapeutic development.

Top