AI Assisted De-Novo Design and Discovery

This article is based on a poster originally authored by Chris King, Chris Murray, Ciaran Doherty, Simon Larsen, Richard Buick, and Adrian Kinkaid.

Background

Antibody Discovery & Machine Learning^1,3

• High-affinity antibodies are traditionally discovered through time-consuming animal immunization and subsequent engineering, such as antibody mutation and selection.
• Machine learning (ML) has significantly advanced the discovery and development of therapeutic antibodies.
• The CDRx^® platform extensively employs ML for database generation, predicting antibody properties and structure, and designing critical CDR loops for binding.

Fig 1. Structural analysis of IgG using in-silico CDRx^® Machine Learning Platform. Image Credit: ELRIG (UK) Ltd.

AI/ML-Ab^TM

• AI can speed up the discovery process by generating and screening antibody fragments and full-length sequences against specific targets for their binding and stability properties.

Methodology – Discovery

• In-silico binding was performed independently of epitope access and antigen conformations, enhancing binder discovery for otherwise inaccessible epitopes.
• Antibodies were engineered based on fragment-variable scaffolds with lead readiness profiles, eliminating the need for numerous iterations of site-specific mutagenesis.
• The crystal structure was not required; libraries of >10,000 sequences can be generated for any number of targets.
• AI-assisted selection plot is provided, showcasing data points for each sequence, including clustered identity and germline information.
• Sequences screened for developability (derived from developable germline families) and patentability (screened against published & patented sequences to reduce freedom to operate limitations).
• The AI mimics human antibody formation and digitizes the process of turning antibodies into drugs.

Fig 2. AI-Assisted Selection Plot. Image Credit: ELRIG (UK) Ltd.

Fig 3. Sequence Identity and Germline Information. Image Credit: ELRIG (UK) Ltd.

Methodology - Engineering

1. Start with an existing antibody sequence that possesses a desirable quality, such as target specificity.
2. Introduce targeted mutations in CDRs to mimic the natural somatic hypermutation process.²
3. Generate a panel or library of mutant antibody sequences and screen for binding affinity.
4. Iterate on sequence optimization by introducing mutations to improve the desired properties.
5. Validate engineered antibodies through rigorous characterization and biochemical assays.
6. Scale up production of selected lead candidate antibodies using appropriate systems.

Fig 4. Fab engineered to possess glycan on CDR. Image Credit: ELRIG (UK) Ltd.

Results

Fig 5. Biolayer Interferometry kinetics analysis of a parental antibody and a variant created using mutations targeting an increase in affinity. Image Credit: ELRIG (UK) Ltd.

Fig 6. Analytical Size Exclusion Chromatography was performed on a parental antibody and a variant created using mutations targeting an increase in monodispersity by way of removal of aggregation-prone motifs. Image Credit: ELRIG (UK) Ltd.

Fig 7. A comparison of binding affinity averages was achieved using phage display (Left) and AI-ML/Ab discovery libraries (Right). Image Credit: ELRIG (UK) Ltd.

Conclusions and Future Work

• The AI approach offers a higher likelihood of observing binders, as sequences are scored based on their evolutionary fitness against our dataset of 2x10⁹ unique human BCR heavy chains.
• De-novo antibody design has a significant impact on the future of discovery.
• Fusion Antibodies offer researchers easy access to an AI pathway to explore the assisted discovery of mAbs, VHH, Bi/Multispecifics, ADCs, and CARs.

Acknowledgments

The team would like to thank the entire Fusion Antibodies team for their contributions to the development of the AI/ML-Ab^TM platform, as well as Prof. Christopher Scott and the Fusion Antibodies Scientific Advisory Panel for their valuable advice. Figures were created using Biorender.com.

References

1. Computational and artificial intelligence-based methods for antibody development (Jisun Kim,1,4 Matthew McFee,2,4 Qiao Fang,2,4 Osama Abdin,2 and Philip M. Kim 1,2,3,*)
2. Kandari, D. and Bhatnagar, R. (2021) Antibody engineering and its therapeutic applications. Int. Rev.
3. Immunol. Published online August 6, 2021.
4. Ganggang Bai, Chuance Sun, Ziang Guo. (2023) Accelerating antibody discovery and design with artificial intelligence.
5. Int. J. Biol. Sci. (2021) Artificial intelligence in the diagnosis of COVID-19: challenges and perspectives.

About ELRIG (UK) Ltd.

The European Laboratory Research & Innovation Group (ELRIG) is a leading European not-for-profit organization that exists to provide outstanding scientific content to the life science community. The foundation of the organization is based on the use and application of automation, robotics and instrumentation in life science laboratories, but over time, we have evolved to respond to the needs of biopharma by developing scientific programmes that focus on cutting-edge research areas that have the potential to revolutionize drug discovery.

Comprised of a global community of over 12,000 life science professionals, participating in our events, whether it be at one of our scientific conferences or one of our networking meetings, will enable any of our community to exchange information, within disciplines and across academic and biopharmaceutical organizations, on an open access basis, as all our events are free-of-charge to attend!

Our Values

Our values are to always ensure the highest quality of content and that content will be made readily accessible to all, and that we will always be an inclusive organization, serving a diverse scientific network. In addition, ELRIG will always be a volunteer led organization, run by and for the life sciences community, on a not-for-profit basis.

Our Purpose

ELRIG is a company whose purpose is to bring the life science and drug discovery communities together to learn, share, connect, innovate and collaborate, on an open access basis. We achieve this through the provision of world class conferences, networking events, webinars and digital content.

Sponsored Content Policy: AZoLifeSciences publishes articles and related content that may be derived from sources where we have existing commercial relationships, provided such content adds value to the core editorial ethos of AZoLifeSciences which is to educate and inform site visitors interested in life science news and information.