EUVIMEDScopri EUVIMED
Open access · CC BY 4.0

Biomedical literature-based clinical phenotype definition discovery using large language models

Samar Binkheder iD, Xiaofu Liu, Michael Wu, Lei Wang iD, Aditi Shendre, Sara K Quinney iD, Wei-Qi Wei, Lang Li iD

Database · 2025-01-18 · 2025

DOI: 10.1093/database/baaf047 ↗

At a glance

Journal article in Database, published online on 2025-09-24. Written by Samar Binkheder and 7 co-authors. Institution: Medical Informatics Unit, Department of Medical Education, College of Medicine, King Saud University, Riyadh 12372, Saudi Arabia. Major MEDLINE topics: Data Mining, Phenotype. Cited 3 times so far (Crossref); 69 references. The full text is freely available (open access, CC BY 4.0). Funded by: NICHD NIH HHS, NLM NIH HHS, NCI NIH HHS.

AbstractSubjectsDati bibliograficiCiteRelated studies

Abstract (original language)

Electronic health record (EHR) phenotyping is a high-demand task because most phenotypes are not usually readily defined. The objective of this study is to develop an effective text-mining approach that automatically extracts clinical phenotype definitions-related sentences from biomedical literature. Abstract-level and full-text sentence-level classifiers were developed for clinical phenotype discovery from PubMed. We compared the performance of the abstract-level classifier on machine learning algorithms: support vector machine (SVM), logistic regression (LR), naïve Bayes, and decision tree. SVM classifier showed the best performance (F-measure = 98%) in identifying clinical phenotype-relevant abstracts. It predicted 459 406 clinical phenotype-related abstracts. For the full-text sentence-level classifier, we compared the performance of SVM, LR, naïve Bayes, decision trees, convolutional neural networks, Bidirectional Encoder Representations from Transformers (BERT), and Bidirectional Encoder Representations from Transformers for Biomedical Text Mining (BioBERT). BioBERT model was the best performer among the full-text sentence-level classifiers (F-measure = 91%). We used these two optimal classifiers for large-scale screening of the PubMed database, starting with abstract retrieval and followed by predicting clinical phenotype-related sentences from full texts. The large-scale screening predicted over two million clinical phenotype-related sentences. Lastly, we developed a knowledgebase using positively predicted sentences, allowing users to query clinical phenotype-related sentences with a phenotype term of interest. The Clinical Phenotype Knowledgebase (CliPheKB) enables users to search for clinical phenotype terms and retrieve sentences related to a specific clinical phenotype of interest (https://cliphekb.shinyapps.io/phenotype-main/). Building upon prior methods, we developed a text mining pipeline to automatically extract clinical phenotype definition-related sentences from the literature. This high-throughput phenotyping approach is generalizable and scalable, and it is complementary to existing EHR phenotyping methods.

Subjects

MeSH terms (MEDLINE)

Data MiningPhenotypeHumansElectronic Health RecordsSupport Vector MachineNatural Language ProcessingPubMedMachine LearningLanguageDecision TreesBayes TheoremAlgorithmsLarge Language Models

Authors and institutions

  • Medical Informatics Unit, Department of Medical Education, College of Medicine, King Saud University, Riyadh 12372, Saudi Arabia.
  • Department of Biomedical Informatics,, College of Medicine, Ohio State University, 1800 Cannon Drive, Columbus, OH 43210, United States.
  • gRED Computational Sciences, Computational Biology & Translation, Genentech, Inc, 1 DNA Way, South San Francisco, CA 94080, United States.
  • Department of Obstetrics and Gynecology, School of Medicine, Indiana University, 950 W Walnut Street, Indianapolis, IN 46202, United States.
  • Department of Biomedical Informatics, Vanderbilt University Medical Center, 2525 West End Ave, Nashville, TN 37203, United States.

ORCID profiles: Samar Binkheder, Lei Wang, Sara K Quinney, Lang Li

Funding

  • NICHD NIH HHS
  • NLM NIH HHS
  • NCI NIH HHS
  • NIA NIH HHS
  • NIGMS NIH HHS

Cite

APA

Binkheder, S., Liu, X., Wu, M., Wang, L., Shendre, A., Quinney, S. K., Wei, W. Q., & Li, L. (2025). Biomedical literature-based clinical phenotype definition discovery using large language models. Database, 2025. https://doi.org/10.1093/database/baaf047

Vancouver

Binkheder S, Liu X, Wu M, Wang L, Shendre A, Quinney SK, et al. Biomedical literature-based clinical phenotype definition discovery using large language models. Database. 2025;2025. doi:10.1093/database/baaf047

BibTeX
@article{Binkheder2025,
  author = {Binkheder, Samar and Liu, Xiaofu and Wu, Michael and Wang, Lei and Shendre, Aditi and Quinney, Sara K and Wei, Wei-Qi and Li, Lang},
  title = {Biomedical literature-based clinical phenotype definition discovery using large language models},
  journal = {Database},
  year = {2025},
  volume = {2025},
  doi = {10.1093/database/baaf047}
}
Start for free →