Identifying Biomedical Entities for Datasets in Scientific Articles: 4-Step Cache-Augmented Generation Approach Using GPT-4o and PubTator 3.0
Bibliographic record
Abstract
Background: The accurate extraction of biomedical entities in scientific articles is essential for effective metadata annotation of research datasets, ensuring data findability, accessibility, interoperability, and reusability in collaborative research. Objective: This study aimed to introduce a novel 4-step cache-augmented generation approach to identify biomedical entities for an automated metadata annotation of datasets, leveraging GPT-4o and PubTator 3.0. Methods: The method integrates four steps: (1) generation of candidate entities using GPT-4o, (2) validation via PubTator 3.0, (3) term extraction based on a metadata schema developed for the specific research area, and (4) a combined evaluation of PubTator-validated and schema-related terms. Applied to 23 articles published in the context of the Collaborative Research Center OncoEscape, the process was validated through supervised, face-to-face interviews with article authors, allowing an assessment of annotation precision using random-effects meta-analysis. Results: The approach yielded a mean of 19.6 schema-related and 6.7 PubTator-validated biomedical entities per article. Within the study's specific context, the overall annotation precision was 98% (95% CI 94%-100%), with most prediction errors concentrated in articles outside the primary basic research domain of the schema. In a subsample (n=20), available supplemental material was included in the prediction process, but it did not improve precision (98%, 95% CI 95%-100%). Moreover, the mean number of schema-related entities was 20.1 (P=.56) and the mean number of PubTator-validated entities was 6.7 (P=.68); these values did not increase with the additional information provided in the supplement. Conclusions: This study highlights the potential of large language model-supported metadata annotation. The findings underscore the practical feasibility of full-text analysis and suggest its potential for integration into routine workflows for biomedical metadata generation.
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 distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
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 teacher head, 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".