Bringing a drug from discovery through clinical trials takes too long and is too expensive, with preclinical costs alone estimated at $15 to $100 million. Employing artificial intelligence (AI) early in the process can lower those costs dramatically.

AI itself isn’t a panacea, though, Jayson Uffens, CTO and chairman of GATC Health, tells GEN. Instead, “Smart computing makes smart people smarter. There’s still a lot of expertise from people on the ground who bring a lot of value—maybe the ultimate value—to the mix.”

GATC Health, an AI-driven therapeutic discovery company, uses AI to raise the floor on opportunities to get high-potential compounds into human studies faster and thereby drive success.

Its proprietary approach to hit and lead identification and program derisking can cut preclinical development costs, according to Uffens, who maintains that the earlier AI is used in a program, the more dramatic the results.

The success GATC Health touts is based on deploying Operon, the company’s proprietary AI platform. Operon deploys in silico models to simulate human biology and takes a multi-omics approach to analysis. That approach has allowed GATC to deliver three to five optimized compounds within six months, claims Uffens, versus the up to 48 months associated with traditional high-throughput screening methods.

Such acceleration occurs by using advanced in silico models to circumvent the “hundreds of thousands of dollars’ worth of experiments performed to get a hit and, ultimately, a lead,” Uffens says.

Rather than relying upon one huge model, he elaborates, “We attack the problem from multiple facets, looking at individual problems with various models and different architectures…and coordinate hundreds of AI models to answer different questions. That’s the starting point. There’s a lot of value in how we curate and parameterize our data in those specific contexts.”

The company also launched the Derisq AI Report, an in-depth analysis of drug candidates that highlights safety concerns, efficacy, and non-obvious risks early, while decision-makers can still modulate those risks.

This predictive intelligence layer is, in fact, a key element of GATC’s clinical trial insurance product. Underwritten by Medical and Commercial International (MCI) under the Lloyd’s of London framework, this insurance product leverages GATC’s predictive capabilities to identify risk. It reimburses the full cost of the trial if safety or efficacy endpoints aren’t met.

Typically, MCI’s preclinical trial insurance clients would provide that company with the relevant trial information, which would be run through the Derisq tool as part of their risk analysis.

Buyers for this insurance tend to be biopharma companies that aren’t large enough to self-insure their own trials. “Capital is expensive for them,” Uffens points out. “The insurance product is there to help them lower the cost of capital and open capital doors that may not be open otherwise.”

Multiomics to Discovery

What’s different about GATC’s approach to AI, Uffens says, is that “We come in, generally, as outsiders.” The founding team includes computer scientists as well as those with strong biology and genetics backgrounds, but not necessarily industry experience.

“We built our technology originally as a genetics interpretation platform,” he recalls, “and expanded it to find additional value.” The company was formed officially in 2020.

The turning point came when GATC became involved in a failed, big pharma program for addiction research.

“(The big pharma company) hadn’t found a solution, but had really valuable data and samples. A partner of ours was working with it to identify biomarkers and thought we could validate them. We discovered that not only could we validate the biomarkers, but we could also identify the therapeutic targets. That’s how we moved from multi-omics analysis into discovery,” Uffens recalls.

Moving forward, “We want to empower researchers,” he says. This means not only helping clients advance existing programs but also by identifying potentially more valuable targets.

Working with GATC

GATC’s key partners most likely will be biotech rather than big pharma, Uffens predicts. And, he notes, “We’re fairly agnostic to therapeutic area.”

“Most of our customers have called us because they want to realize the benefits of AI sooner rather than later,” Uffens says. “There is a lot of risk in the space. Folks who are willing to adopt AI at this stage…are looking for additional help before they risk more capital…” to solve particular challenges.

For a company to begin working with GATC, he explains, “The data we’re looking for is very similar to what they would include in an Investigational New Drug (IND) package. The earlier they are in the process, the less data they will have, but, at a minimum, we need some particulars on their therapeutic’s chemistry and the intended mechanism of action.”

Challenges

Drug development is a difficult space with plentiful challenges, he admits. Therefore, “We approach things as a tech company. We iterate through a problem and find where we can succeed or fail as quickly as we can to develop a solution. We’ve gone through multiple generations of architectures, finding ways that work best.”

The next milestone is to accumulate multiple successes with Operon and Derisq in human trials. “‘Wins in humans’ is our [next] frontier,” he says. That includes wins for its insurance underwriting partners as well as for companies working directly with GATC to advance therapeutics to human trials.

As part of that goal, GATC and BioAtla are closing a deal for a Phase III trial of ozuriftamab vedotin for oropharyngeal squamous cell carcinoma and to further develop conditionally active biologic senolytic therapies. Termed a special purpose vehicle transaction—a financial entity designed to hold specific assets that last for the life of the project—the $40 million deal formed Inversagen AI, LLC, to leverage the strengths of the founding companies.

“GATC and BioAtla are equal partners in Inversagen,” Uffens says. “GATC will own a percentage of ozuriftamab vedotin and a larger stake in future joint discoveries,” thus potentially discovering new therapeutic combinations that may be effective as conditionally active biologics.

Currently, the GATC is fine-tuning its own project prioritization. “The AI landscape is both beneficial and challenging,” Uffens acknowledges. “People have certain expectations about what AI can and should do, how it works, and how they might adopt it. Getting them to hear our unique perspective comes back to our focus on wins in humans.”

The post AI <i>In Silico</i> Multi-Omics Technique Cuts Therapeutic Development Costs appeared first on GEN – Genetic Engineering and Biotechnology News.