Evaluating ChatGPT as an Agent for Providing Genetic Education
Walton, N.; Graceffo, S.; Sutherland, N.; Kozel, B.; Danford, C.; McGrath, S.
Show abstract
Genetic disorders are complex and can greatly impact an individuals health and well-being. In this study, we assess the ability of ChatGPT, a language model developed by OpenAI, to answer questions related to three specific genetic disorders: BRCA1, MLH1, and HFE. ChatGPT has shown it can supply articulate answers to a wide spectrum of questions. However, its ability to answer questions related to genetic disorders has yet to be evaluated. The aim of this study is to perform both quantitative and qualitative assessments of ChatGPTs performance in this area. The ability of ChatGPT to provide accurate and useful information to patients was assessed by genetic experts. Here we show that ChatGPT answered 64.7% of the 68 genetic questions asked and was able to respond coherently to complex questions related to the three genes/conditions. Our results reveal that ChatGPT can provide valuable information to individuals seeking information about genetic disorders, however, it still has some limitations and inaccuracies, particularly in understanding human inheritance patterns. The results of this study have implications for both genomics and medicine and can inform future developments in this area. AI platforms, like ChatGPT, have significant potential in the field of genomics. As these technologies become integrated into consumer-facing products, appropriate oversight is required to ensure accurate and safe delivery of medical information. With such oversight and training specifically for genetic information, these platforms could have the potential to augment some clinical interactions.
Matching journals
The top 6 journals account for 50% of the predicted probability mass.
Similar papers in this journal
- GUIA: a digital platform to facilitate result disclosure in genetic counseling 94%
- Development and evaluation of a novel educational program for providers on the use of polygenic risk scores 94%
- ClinGen Variant Curation Interface Workshops: Training Variant Scientists on an International Platform 93%
Similar papers in this journal
- Analysis of large-language model versus human performance for genetics questions 96%
- Defining the Critical Educational Components of Informed Consent for Genetic Testing: Views of US-Based Genetic Counselors and Medical Geneticists 96%
- A Qualitative Study Exploring the Consumer Experience of Receiving Self-Initiated Polygenic Risk Scores from a Third-Party Website 93%
Similar papers in this journal
- ChatGPT- versus human-generated answers to frequently asked questions about diabetes: a Turing test-inspired survey among employees of a Danish diabetes center 92%
- Clinical code sets and the problem of redundancy in code set repositories 92%
- Protocol For Human Evaluation of Artificial Intelligence Chatbots in Clinical Consultations 91%
Similar papers in this journal
- A Pooled Electronic Consultation Program to Improve Access to Genetics Specialists 94%
- Evaluation of DNA-poli: study protocol of a randomised controlled trial to assess a digital platform for family cascade genetic testing and predictive genetic counselling 92%
- Early Diagnosis of Vascular Ehlers-Danlos Syndrome Through AI-Powered Facial Analysis: Results from the Montalcino Aortic Consortium 89%
Similar papers in this journal
- Design and user experience testing of a polygenic score report: a qualitative study of prospective users 94%
- genepanel.iobio - an easy to use web tool for generating disease- and phenotype-associated gene lists 93%
- GeneTerpret: a customizable multilayer approach to genomic variant prioritization and interpretation 91%
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.