Health AI Central
Peer-Reviewed Publication
Nucleic Acids Res2026;54(14)July 17, 2026Journal Article

Systematic contextual biases in SegmentNT potentially relevant to other nucleotide transformer models.

Mark T W Ebbert1,2,3, Anna Ho2,4, Madeline L Page1,2,3, Bram Dutch5, Blake K Byer2,4, Kristen L Hankins6, Hady Sabra1,6,7, Bernardo Aguzzoli Heberle1,3, Mark E Wadsworth1,2,3, Grant A Fox1,3, Bikram Karki2,5, Caylin Hickey4,6, David W Fardo1,8, Cody Bumgardner2,4,6, Yasminka A Jakubek1,2,9, Cody J Steely1,2,9, Justin B Miller1,2,4,7
1Sanders-Brown Center on Aging, University of Kentucky, Lexington, KY 40506, United States.
2Division of Biomedical Informatics, Department of Internal Medicine, University of Kentucky, Lexington, KY 40506, United States.
3Department of Neuroscience, University of Kentucky, Lexington, KY 40506, United States.
4Department of Pathology and Laboratory Medicine, University of Kentucky, Lexington, KY 40506, United States.
5Department of Computer Science, University of Kentucky, Lexington, KY 40506, United States.
6Institute for Biomedical Informatics and the Center for Applied Artificial Intelligence, College of Medicine, University of Kentucky, Lexington, KY 40506, United States.
7Department of Microbiology, Immunology, and Molecular Genetics, University of Kentucky, Lexington, KY 40506, United States.
8Department of Biostatistics, University of Kentucky, Lexington, KY 40506, United States.
9Markey Cancer Center, University of Kentucky, Lexington, KY 40506, United States.
PubMed DOI

Abstract

Recent advances in large language models have extended to genomic applications, yet model robustness relative to context is unclear. Here, we demonstrate two intrinsic biases (input sequence length and nucleotide position) affecting SegmentNT results, a model included with the Nucleotide Transformer that provides nucleotide-level predictions of biological features. We demonstrate that nucleotide p…

Create a free account to keep reading

Free members get 10 full research views every month across publications, clinical trials, FDA clearances, adverse events, and NIH grants. No credit card required.

Create Free AccountSign In

Want unlimited research access? See Pro plans

Data Accuracy Notice: Research intelligence on Health AI Central is aggregated from public sources (PubMed, ClinicalTrials.gov, FDA, NIH, CMS, and others) and refreshed nightly. Classifications and derived metrics are produced by automated methods described in our Methodology. We recommend verifying critical data points against the primary sources before making decisions.