PDAO™ AI | TGF-β Corpus | Oncotelic (Preview)
PDAO™ AI Evidence interrogation • auditable
Let the dataset speak • hypothesis-first discovery
✨ Built for difficult-to-treat cancers

AI that interrogates evidence—
not a black-box predictor.

PDAO™ AI is Oncotelic’s evidence-interrogation platform. Instead of training a bespoke model to imitate your dataset, we structure, embed, cluster, and query large bodies of biomedical knowledge so the dataset can speak—yielding reproducible, auditable, testable hypotheses.

Discord access: TGF-β corpus community channel • Invite: https://discord.gg/Rj5PaHUfzw

Interrogation layer (no bespoke model training)

We prioritize evidence organization and retrieval over training a custom black-box predictor. The goal: surfaced signals you can validate.

Reproducible and auditable

Semantic retrieval + clustering + query traces provide a clear chain from question → evidence → hypothesis.

Translation-oriented outputs

PDAO™ AI generates hypothesis candidates: pathway dependencies, tumor contexts, biomarker patterns linked to endpoints.

How it works (at a glance)

1

Ingest literature + structured datasets.

2

Embed into a semantic space (similarity).

3

Cluster recurring themes to expose patterns.

4

Query reproducibly to generate hypotheses.

Output is evidence-linked hypotheses—designed to be testable against biology and outcomes.

TGF-β Corpus: comprehensive, conversant knowledge

Oncotelic has curated a comprehensive TGF-β literature corpus containing 125,000+ PubMed abstracts—representing the totality of scientific knowledge related to TGF-β across oncology, immunology, fibrosis, metabolism, and translational therapeutics.

The corpus is embedded, clustered, and indexed in PDAO™ AI so researchers can interrogate the evidence to generate new hypotheses and new insights.

What you get: interactive discovery (semantic search + clustered themes) rather than a static bibliography.

What researchers can do

A

Hypothesis mining: identify recurring mechanistic links around endpoints (survival, resistance, immune escape).

B

Context mapping: discover tumor microenvironment contexts that repeatedly co-occur with TGF-β signals.

C

Cross-domain synthesis: connect immunology, fibrosis, metabolism, and oncology without being trapped by one training set.

D

Evidence traceability: maintain a chain from question → retrieved evidence → proposed hypothesis.

PDAO™ AI → TGF-β Corpus → Hypothesis → Translation

A simple schematic you can reuse in decks and web collateral.
PDAO™ AI Evidence interrogation layer Ingest • Embed • Cluster • Query Audit trail from question → evidence Reduce training-set bias TGF-β Corpus 125K+ PubMed abstracts Semantic index + clustered themes Totality of TGF-β knowledge Cancer • Immunology • Fibrosis • Metabolism Outputs Testable hypotheses Signals recurring across contexts Pathways • contexts • biomarkers Evidence-linked, not black-box scores Translation Actionability Biomarkers • combos • trials Prioritize what to validate next Accelerate decisions in hard cancers Oncotelic Therapeutics • PDAO™ AI • Let the dataset speak

Discord access

The TGF-β corpus community is hosted on Discord to support collaborative exploration, hypothesis discussion, and onboarding.

https://discord.gg/Rj5PaHUfzw

Onboarding guide (researchers)

A lightweight path to get value in 15 minutes once you join Discord.

1

Join Discord and read the welcome + rules.

2

Introduce yourself: tumor area, focus (TME, fibrosis, immunology), and what you want to discover.

3

Start with 3 queries: (i) TGFB2 + survival, (ii) TGF-β + macrophages, (iii) TGF-β + resistance.

4

Capture outputs: cluster summaries + top abstracts + your testable hypothesis statement.

5

Share & iterate in the hypotheses channel: what recurs, what contradicts, and what to validate next.

Suggested format: Question → Top evidence → Cluster themes → Hypothesis → Validation plan

Generate a copyable access request

Use this to route inbound interest internally (research, collaboration, investor diligence).