Oncotelic Unlocks AI Platform and Cancer Data Trove for Researchers

AGOURA HILLS, CA – December 22, 2025 – Oncotelic Therapeutics, an RNA-based therapeutics company, has announced a significant move to accelerate biomedical discovery by opening access to its proprietary PDAOAI platform. In a parallel effort to foster collaboration, the company is also providing the global research community with its entire curated knowledge corpus on TGF-β, a key signaling protein implicated in numerous diseases, including cancer. This comprises over 125,000 scientific abstracts.

The initiative invites qualified researchers, academic institutions, and industry partners to engage directly with these powerful tools through a dedicated Discord research channel, creating a novel, interactive ecosystem for hypothesis generation and collaborative science. This decision marks a pivotal moment in the company's strategy, shifting from purely internal use to democratizing a tool that has been instrumental in its own clinical and preclinical programs.

A New Philosophy for AI in Biomedicine

In an era where artificial intelligence is often synonymous with large language models (LLMs), Oncotelic is championing a different approach. The PDAOAI platform is not designed to be trained on proprietary data to make opaque predictions. Instead, it functions as a sophisticated "interrogation layer," built to navigate the overwhelming sea of publicly available biomedical data and extract meaningful, verifiable signals.

The platform works by ingesting vast quantities of information—from peer-reviewed literature to structured clinical datasets—and embedding the content into a semantic space. It then applies advanced clustering algorithms to identify recurring biological themes and enables researchers to perform structured, repeatable queries. This process is designed to surface patterns that are already present in the data, allowing them to emerge organically.

“Our philosophy is simple: let the dataset speak,” stated Vuong Trieu, co-author and executive contributor. “In oncology today, the constraint is no longer data access—it is signal discovery and navigation. Over-training models on narrow datasets introduce bias and over-training models on large datasets dumb down the model due to noises. PDAOAI is built to surface reproducible, citation-backed workflows that generate testable hypotheses rather than opaque, black-box predictions.”

This methodology directly confronts a core challenge in biomedical AI: the risk of creating "black-box" systems whose conclusions are difficult to trace and validate. By ensuring that every insight is linked back to its source literature or data point, PDAOAI empowers researchers to build upon a foundation of existing evidence, fostering trust and reproducibility. The platform has already had a transformative effect internally.

“PDAOAI has enabled our researchers to elevate to a level that was not possible before. It has completely changed the game and has made research both more in depth and efficient,” commented Scott Myers, Product Manager at Oncotelic.

From Theory to Practice: Peer-Reviewed Validation

The power of the PDAOAI platform is not merely theoretical. It has served as a crucial engine for evidence synthesis in studies that have culminated in seven peer-reviewed publications, demonstrating its real-world utility. These papers, published in prestigious journals such as the International Journal of Molecular Sciences and Cancers, showcase the platform's ability to contextualize complex biological axes associated with patient survival across different tumor types and microenvironments.

A compelling example of the "dataset speaking" paradigm is the repeated emergence of TGFB2 as a central, survival-associated axis across multiple cancers. Through PDAOAI, researchers identified consistent patterns linking TGFB2 to patient outcomes in diverse malignancies, including:

• Pancreatic Ductal Adenocarcinoma (PDAC): Studies highlighted how TGFB2 expression and methylation status can predict overall survival. One paper in the International Journal of Molecular Sciences detailed how TGFB2 gene methylation in tumors with low CD8+ T-cell infiltration drives positive prognostic outcomes in PDAC patients.
• Glioblastoma: A study in Cancers revealed the positive prognostic survival impacts of methylated TGFB2 and MGMT in adult glioblastoma patients.
• Ovarian and Breast Cancers: Other research used the platform to identify potential TGFB2-dependent and independent prognostic biomarkers for patients treated with the chemotherapy drug Taxol.

While no single gene can explain the complexity of cancer, the consistent recurrence of TGFB2 as a critical node provides a high-confidence anchor for developing downstream biomarker strategies and therapeutic combinations. This is particularly relevant for Oncotelic, as this AI-driven insight directly informs and validates the clinical strategy for its lead asset, OT-101 (Trabedersen), an antisense agent designed to target TGFB2.

Unlocking a Universe of TGF-β Knowledge

Central to this initiative is the opening of Oncotelic's comprehensive TGF-β literature corpus. Transforming growth factor-beta (TGF-β) is a family of proteins that plays a crucial role in cell proliferation, differentiation, and death. While essential for normal biological function, its dysregulation is a hallmark of many diseases, driving fibrosis, suppressing immune responses in cancer, and influencing metabolic disorders.

The corpus, containing more than 125,000 PubMed abstracts, represents the near-totality of published scientific knowledge on the subject. By integrating this vast library into the PDAOAI environment, Oncotelic has turned a static collection of references into a dynamic, interactive knowledge base.

Researchers accessing the platform can now "converse" with this dataset, using it to:
* Generate novel hypotheses about TGF-β's role in specific disease contexts.
* Discover non-obvious connections between TGF-β biology and other pathways.
* Identify overlooked or non-canonical mechanisms that could lead to new therapeutic targets.
* Develop fresh mechanistic and translational insights to guide their own research.

This open-access model is designed to spur collaborative exploration and accelerate progress across the entire field of TGF-β-driven science, moving beyond the confines of a single organization's R&D efforts.

A Collaborative Future for Drug Discovery

Oncotelic's strategic decision to open its platform and data is a clear call for a more collaborative approach to drug discovery. The dedicated Discord channel serves as the central hub for this new community, providing an interactive forum for hypothesis exploration, signal interrogation, and discussion among leading minds in academia and industry.

Internally, PDAOAI already serves as a core discovery and strategy engine, guiding biomarker identification, precision trial design, and the interpretation of clinical endpoints. The company emphasizes that the platform is designed to augment expert judgment, not replace it. By systematically and transparently surfacing signals that warrant focused experimental validation, PDAOAI allows scientists to spend less time sifting through data and more time designing the critical experiments that lead to breakthroughs.

By extending this capability to the broader scientific community, Oncotelic is positioning itself as a central player in a new, data-driven ecosystem. This move not only highlights the maturity and confidence in its proprietary technology but also aligns with a growing movement toward open science, recognizing that the monumental challenges of diseases like cancer are best tackled through collective intelligence. Researchers and organizations interested in partnership opportunities or engaging with the platform are encouraged to join the community and explore the future of evidence-based discovery.