FRONTEO Confirms the Cell Proliferation - Inhibitory Effect of Novel Pancreatic Cancer Target Candidates, Identified by its AI Drug Discovery Support Service "Drug Discovery AI Factory"

Tokyo, Japan, July 23, 2025 - FRONTEO, Inc. announced that, using its AI drug discovery support service Drug Discovery AI Factory (hereinafter “DDAIF”) ¹, it has identified candidate drug discovery target molecules² for pancreatic cancer, and confirmed that they inhibit cancer cell proliferation in vitro (test‑tube) experiments³. This experiment, conducted by FRONTEO, aimed to demonstrate the effectiveness of DDAIF that identifies correlations between target molecules and diseases not reported in the literature.

Shortening target identification from about two years to two days; successfully pinpointed numerous highly novel candidate molecules Using DDAIF, FRONTEO identified 17 candidate target genes from approximately 20,000 human genes in two days. In the conventional drug discovery process, target identification often takes more than two years, and finding highly novel targets not reported in the literature is considered particularly challenging, regardless of the time invested. DDAIF’s ability to accelerate this critical stage of drug discovery— “target discovery.”

Of the 17 genes identified, six were found in experiments to inhibit pancreatic cancer cell proliferation, four had no reported association with pancreatic cancer in any published paper, while the remaining two had only a single report each (as of April 19, 2025)⁴—making them novel targets.

Pancreatic cancer: a disease with an exceptionally high unmet medical need
Pancreatic cancer has the lowest five-year survival rate of all major cancers — less than 10%⁵. Treatment options remain limited, relying heavily on highly toxic chemotherapeutic agents. Once chemotherapy becomes ineffective, alternative treatments virtually nonexistent, making pancreatic cancer one of the diseases with the greatest unmet medical need⁶.

Extraction of numerous highly novel target molecule candidates
Of the six genes confirmed to inhibit cell proliferation, four had no reported association with pancreatic cancer in any published paper. For the remaining two, only one paper existed for each. Thus, a large number of highly novel target molecule candidates were identified.

Significant inhibitory effects on cell proliferation
When the function of the 17 genes was suppressed in pancreatic cancer cell lines, six genes showed approximately 40–60% inhibition of cancer cell proliferation compared to untreated controls —an effect comparable to that of KRAS, a gene known to play a major role in promoting carcinogenesis. These findings indicate that these six genes have strong potential as drug targets.

Going forward, we plan to collaborate with co-creation partners, including Institute of Science Tokyo, to clarify the mechanism of action of the target molecule candidates that have shown efficacy and to verify them through animal experiments. In the future, we will also consider out-licensing⁷ them to pharmaceutical companies.

Comment from Masayuki Murata, Specially Appointed Professor, Institute of Integrated Research, Institute of Science Tokyo
"The ability to narrow down unreported drug targets to several candidates in a short time—even within days depending on the analysis target—is an innovation unimaginable with conventional approaches. Although this verification was conducted at the cellular level, it clearly demonstrates how KIBIT can greatly contribute to accelerate drug discovery, and improve its success rate.
At Institute of Science Tokyo, our research group seeks to deepen understanding of the life phenomena underlying drug discovery and cell‑based therapies. We are developing a variety of cell function analysis and evaluation methods. Through our ongoing joint research with FRONTEO, we will systematically perform functional verification in cell‑based experiments on the identified target molecule candidates, with the expectation of accelerating drug discovery by pinpointing molecules with proven functional activity.

Challenges in drug discovery
The establishment of the Public‑Private Council for Enhancing Drug Discovery Capability by the Cabinet Office of Japan in June 2025⁸ underscores that strengthening drug discovery capability has become an urgent national priority. Japan has long faced a persistent pharmaceutical trade deficit⁹, and building a system that does not rely on overseas pharmaceutical supply is now a pressing need.

However, drug discovery research has already reached a mature stage. Many of the easy‑to‑develop drugs are already on the market, while the growing complexity and diversity of diseases have made new target discovery increasingly difficult. The resulting “depletion” of target molecules has become a serious issue.

Internationally, there is a trend toward limiting the approval of generic drugs. As a result, the development of Innovative drugs with a strong first‑mover advantage, known as “first‑in‑class” new drugs¹⁰, has become essential for both corporate competitiveness and national pharmaceutical strategies. Achieving this requires discovering “unknown target molecules” not described in any papers or databases, yet identifying such information has been extremely difficult with conventional methods, including traditional AI approaches.

Significance of this project using DDAIF
DDAIF tackles these structural challenges by applying its proprietary natural language processing AI to identify “unknown target molecules” not reported in any literature, thereby enabling innovative drug discovery support.

FRONTEO will continue contributing to solving structural challenges in life sciences and sustainable development of the medical industry through R&D of natural language processing‑specialized AI and its social implementation.

■Notes

1. An AI drug discovery support service where FRONTEO’s AI drug discovery experts utilize the natural language processing technology of the company’s proprietary AI KIBIT and unique analytical methods to provide target molecule and indication discovery and supporting hypotheses.
2. Molecules (genes) that are targeted by drugs.
3. It was confirmed that the target molecule candidates identified by DDAIF had a statistically significant inhibitory effect on pancreatic cancer cell proliferation by temporarily suppressing their functions.
4. PubMed AND search of gene names and “pancreatic cancer” in paper titles and abstracts (as of April 19, 2025).
5. National Cancer Center Japan: Latest Cancer Statistics.
6. Unmet medical needs: diseases for which no effective treatment exists.
7. In the pharmaceutical industry, companies may license out development or sales rights of drug candidates to others—licensing out refers to granting such rights to pharmaceutical companies.
8. Cabinet Office: Vision Conference for Delivering the Latest Pharmaceuticals to the People Quickly by Improving Drug Discovery Capability.
9. Cabinet Office: Document 5: Current Status of Global and Japanese Drug Discovery.
10. The first drug approved in that category of pharmaceuticals (first‑in‑class).
    
    

About FRONTEO Drug Discovery AI Factory (DDAIF)
Website: AI Drug Discovery Support Platform “Drug Discovery AI Factory” - FRONTEO LifeScience AI

FRONTEO Drug Discovery AI Factory (DDAIF) is an AI drug discovery support service that combines KIBIT, an AI specialized in natural language processing (patented in Japan and the U.S.), with the expertise of FRONTEO's drug discovery researchers and AI engineers. By analyzing disease-related gene networks and building hypotheses about target candidates, DDAIF provides powerful support for researchers' decision-making in drug development.
This service has already been adopted by several major pharmaceutical companies and has a proven track record.

■About FRONTEO, Inc. https://www.fronteo.com/en/

FRONTEO provides its proprietary specialized AI“KIBIT” to support expert judgment across domains confronting social challenges. Its unique natural language processing technology (patented in Japan, the U.S., and Europe) enables fast, high precision analysis without reliance on training data volume or computational power. Additionally, patented technology that maps (visualizes structure) analyzed information allows KIBIT to directly influence expert insights, and in recent years, KIBIT has been applied in hypothesis generation and target discovery for drug development.

To achieve the philosophy of "providing solutions that do not overlook risks and opportunities buried in records and realizing fairness in the information society," FRONTEO is promoting the social implementation of KIBIT's unique technology and AI solutions in the following areas: life science AI, risk management (business intelligence and compliance support, economic security, and legal tech AI), and DX (business intelligence and professional support)

*The technology used in Drug Discovery AI Factory is protected by 21 patents held by FRONTEO in Japan, South Korea, the U.S., and Europe.
**FRONTEO, KIBIT, and Drug Discovery AI Factory are trademarks or registered trademarks of FRONTEO in Japan, Europe, the U.S., and South Korea.