Ordaōs/NonExomics Deal Shines Light on Novel Mini-Proteins


Ordaōs and NonExomics announced a research pact Wednesday to develop mini-proteins, called miniPROs, for three specific difficult-to-target rare cancers.

In an interview with BioSpace, Ordaōs CEO David Longo indicated the research collaboration’s focus is on rare cancers, that the commonality is “difficult targets” and that the miniPROs™ are “uniquely capable of being inserted into the tumor microenvironments.”

David Longo_Ordaos Under the partnership, Ordaōs will use its Design Engine to identify novel miniPRO candidates to meet specifications set out by NonExomics. NonExomics will then conduct preclinical studies to determine if they meet the product profiles.

Ordaōs focuses on designing novel, proprietary miniPRO mini-proteins. It uses the Ordaōs Design Engine, which leverages “continuous learning loops and proprietary datasets to translate human-targeted product criteria into machine-designed mini-proteins.”

This technology platform starts with amino acids, then generates, analyzes and ranks billions of protein sequences and hundreds of thousands of protein structures and known properties in order to design miniPRO proteins.

The company then evaluates them in assays for various qualities, including protein structure, binding specificity and affinity. As a result, Ordaōs believes these proteins will have more focused efficacy with fewer adverse side effects while also being easier and less expensive to develop than traditional proteins.

Longo described the miniPROs as “40 to 160 amino acid proteins,” which are larger than a peptide. Because of their small size, they can be inserted into the tumor microenvironment and can be “just as efficacious as antibodies … but smaller.”

One of the differentiators for The Ordaōs Design Engine is “We’re starting from nothing and building amino acid by amino acid,” which provides a “higher order, holistic view of the protein structure," Longo said. In short, the AI platform “learns how to learn, and learns how to learn better.”

He said the ranking system could optimize across all properties simultaneously. It allows the company to maximize the speed of delivery of its drug candidates and its novelty and probability of clinical success.

In late August, Ordaōs completed a $5 million seed financing in support of its technology platform.

NonExomics’ approach leverages artificial intelligence (AI) and proteogenomics - how DNA is linked to the proteins it codes for - to identify hundreds of new disease-associated biomarkers that have been tied to more than 1,365 diseases.

In June, NonExomics launched the new technology platform with the goal of identifying whether these novel proteins can act as drug targets for therapeutic benefit or potential diagnostic methods.

To date, it has validated the transcript and protein expression of almost 248,135 non-exomic regions, whose mutations are tied to a variety of diseases. This includes cancer, schizophrenia, ALS and age-related macular degeneration.

NonExomics is one of seven genomics startups picked by the Illumina Accelerator, which included seed investment, access to Illumina gene sequencers, as well as business advice, genomics expertise and lab space next to its campuses in Cambridge, U.K. or the San Francisco Bay area.

NonExomics was co-founded by Sudhakaran Prabakaran, Ph.D., and Ruchi Chauhan, Ph.D., in 2016, both from Harvard and Cambridge Universities.