Building a literature knowledge base towards transparent biomedical AI
Bibliographic record
Abstract
Abstract As artificial intelligence (AI) continues to advance and scale up in biomedical research, concerns about AI’s trustworthiness and transparency have grown. There is a critical need to systematically bring accurate and relevant biomedical knowledge into AI applications for transparency and provenance. Knowledge graphs have emerged as a powerful tool that integrates heterogeneous knowledge by explicitly describing biomedical knowledge as entities and relationships between entities. However, PubMed, the largest and most comprehensive repository of biomedical knowledge, exists primarily as unstructured text and is under utilized for advanced machine learning tasks. To address the challenge, we developed LiteralGraph, a computational framework to extract biomedical terms and relationships from PubMed literature into a unified knowledge graph. Using this framework, we established the Genomic Literature Knowledge Base (GLKB), which consolidates 14,634,427 biomedical relationships between 3,276,336 biomedical terms from over 33 million PubMed abstracts and nine well-established biomedical repositories. The database is coupled with RESTful APIs and a user-friendly web interface that makes it accessible to researchers for various usages. We demonstrated the broad utility of GLKB towards transparent AI in three distinct application scenarios. In the LLM grounding scenario, we developed a Retrieval Augmented Generation (RAG) agent to reduce LLM hallucination in biomedical question answering. In the hypothesis generation scenario, we elucidated the potential functions of RFX6 in type 2 diabetes (T2D) using the vast evidence from PubMed articles. In the machine learning scenario, we utilized GLKB to provide semantic knowledge in predictive tasks and scientific fact-checking.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.007 | 0.041 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.002 |
| Bibliometrics | 0.027 | 0.011 |
| Science and technology studies | 0.002 | 0.001 |
| Scholarly communication | 0.006 | 0.008 |
| Open science | 0.004 | 0.009 |
| Research integrity | 0.002 | 0.003 |
| Insufficient payload (model declined to judge) | 0.009 | 0.007 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".