“Our mission at OmniAb is to enable the rapid development of innovative therapeutics by pushing the frontiers of drug discovery technologies – and we do that through a blending of our in silico, in vitro, and in vivo competencies. OmniDeep^™ is a suite of in silico, AI and machine learning tools that we leverage to realize the full potential of the naturally-optimized immune repertoires produced in our transgenic animals. This integration enables us to deeply mine a vast antibody sequence space aided by in vitro phenotypic screening and next generation sequencing data, and our technologies are designed to discover best-in-class antibody candidates with our industry and academic partners.”

Antibodies are a rapidly growing class of drugs used in high-value and high-impact therapeutic areas like oncology, autoimmunity, inflammation, infectious diseases, and more. When developed skillfully, antibodies can exhibit exquisite specificity and high on-target binding compared to traditional, small-molecule drugs. They can also persist in the blood, leading to improved efficacy and less frequent dosing. These advantages have led to a significant investment in developing antibodies across a broad range of therapeutic areas.

However, the road to clinical testing is filled with pitfalls and bottlenecks, requiring innovative, streamlined, and cost-effective solutions to generate developable therapeutic antibody candidates. In the discovery stage, a pool of potential antibodies specific for a target of interest can be generated using animal-based techniques (e.g., in vivo discovery), tools outside of the context of a living organism (e.g., in vitro discovery), or computational tools (e.g., in silico discovery). The pool of generated antibodies is triaged through a screening and characterization cascade to converge upon a few antibody candidates that show favorable binding profiles, promising efficacy in treating a target disease, good manufacturability, and a low risk of toxicity to humans. Success depends on having robust discovery tools and experience across many scientific and technological disciplines to navigate the challenges that can arise along the way.

In Silico Solutions

The combination of in vivo systems with in silico tools can be especially powerful. Artificial intelligence (AI) and machine learning (ML) are relatively new tools for accelerating antibody discovery and development by streamlining the ‘funnel’ of plausible solutions. AI and ML can facilitate rapid data mining and analysis of vast data sets, efficacy prediction, and structural modeling, all of which further enable decision-making and potentially provide non-obvious insights during the antibody discovery and development process.

OmniAb, Inc., an antibody discovery technology company  with a proven track record of generating antibodies that not only enter the clinic but achieve market approval to treat patients, harnesses the drug discovery capabilities of AI, ML, deep learning, and a suite of other computational tools. They are using these in silico tools to navigate massive data sets to identify and prioritize candidate antibodies that meet a partner’s desired target product profile from the billions of possibilities produced from their naturally optimized immune repertoires. The company’s in silico platform, called OmniDeep^™, is a recent addition to its expanding technology stack, enabling partners to find the antibody candidates from OmniAb’s suite of animal platforms that best meet their criteria.

Using Biological Intelligence^™ for Hard-to-Drug Targets

OmniAb operates under the data-backed assertion that in vivo systems are imbued with a natural ability to generate high-quality antibodies. During animal-based antibody generation efforts, an antigen (a foreign molecule that activates the immune system and leads to the production of antibodies) is presented to an animal through immunization. The resulting antibodies are optimized in vivo through an iterative process that preferentially selects molecules with excellent specificity, stability, and other “developability” characteristics, which ultimately can help to provide patients with safe and efficacious treatments.

Over the past decade, OmniAb has engineered several animal platforms for this type of antibody discovery, including mice, rats, and chickens that express fully human antibodies. The main benefit of these technologies is that they generate naturally optimized antibody sequences without the time, challenges, and pitfalls of a humanization process.

According to OmniAb, the ability of the immune system in OmniAb’s engineered animals to create optimized antibodies for human therapeutics—called Biological Intelligence (BI)—increases the efficiency and probability of success in therapeutic antibody discovery. But for many drug developers, antigen design can undercut this advantage and be a significant challenge, particularly for certain classes of targets, such as ion channels. These are tempting targets for drug development due to their involvement in neurological disorders, yet they have a complex structure and are difficult to purify. OmniAb’s advanced immunization strategies, including mRNA-lipid nanoparticle (LNP) formulations and custom protein preparations, provide the broader drug discovery industry access to differentiated tools for developing antibodies against these hard-to-drug targets.

Integrating AI with BI and Beyond, Across OmniAb’s Entire Technology Stack

Generating the best and most diverse antibody repertoires is just one part of the discovery process. Drug developers must then identify, screen, and characterize them.

OmniAb excels in these areas by providing comprehensive services that integrate OmniDeep to help select the most promising therapeutic candidates. At the heart of OmniAb’s screening and characterization technology stack is the xPloration^® platform, a high-throughput instrument that uses ~1.5 million capillaries to separate individual antibody-producing cells and is capable of collecting multi-parameter data on the functional activity of antibody candidates. The most promising antibody hits can be further analyzed through DNA sequencing, allowing OmniAb’s partners to optimize a promising candidate’s specificity, binding strength, or developability.

OmniDeep can use this sequence and functional data to predict the binding strength of an antibody candidate to a drug target. It can also use structure-based modeling and deep learning to filter out candidates for developability problems and make predictions about antibody candidates that have not been functionally tested. This analysis provides drug developers with a rapid way to explore the potential of other candidates without putting time and money into additional testing. The synergy between OmniDeep, xPloration, and OmniAb’s BI tools empowers the company’s partners to make more informed, data-driven decisions about which antibody candidates have the best chance of clinical success.

An Ever-Evolving Industry Requires an Ever-Evolving Partner

OmniAb’s AI- and ML-powered, end-to-end antibody discovery and development platform has attracted fruitful partnerships with pharmaceutical giants and academic institutions. Many of these partnerships have led to the regulatory approval of therapeutic antibodies. OmniAb’s technology has empowered the company’s partners to successfully develop antibodies approved in the US, Europe, and Asia.

As investment in the antibody market continues and more and more antibodies enter clinical and pre-clinical testing, the industry will always look for ways to accelerate and innovate. AI and other computational methods are the latest in this effort, and their power is undeniable. OmniAb is at the forefront of this movement, adopting technology at the leading edge of the industry. The company’s approach enables them to push antibodies into new therapeutic areas for their drug development partners looking to bring better solutions to patients.