EUVIMEDEUVIMED entdecken
Crossref

Knowledge-Driven Feature Selection and Engineering for Genotype Data with Large Language Models.

Li, Dawei, Duong-Tran, Duy, Zixuan Wen, Liu, Xiaoxi, Yang, Shu, Lee, Joseph, Tianlong Chen und 5 weitere

Carolina Digital Repository (University of North Carolina at Chapel Hill) · 20.06.2025

DOI: 10.17615/fjd2-j454 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Carolina Digital Repository (University of North Carolina at Chapel Hill), erschienen am 20.06.2025. Verfasst von Li und 16 weiteren Autor:innen.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Predicting phenotypes with complex genetic bases based on a small, interpretable set of variant features remains a challenging task. Conventionally, data-driven approaches are used for this task, yet the high dimensional nature of genotype data makes the analysis and prediction difficult. Motivated by the biomedical knowledge encoded in pre-trained LLMs and the emerging applications for genetics, we set to examine the ability of LLMs in feature selection and engineering for tabular genotype data, with a novel knowledge-driven framework. We develop FREEFORM, Free-flow Reasoning and Ensembling for Enhanced Feature Output and Robust Modeling, designed with chain-of-thought and ensembling principles, to select and engineer features with the intrinsic knowledge of LLMs. Evaluated on two distinct genotype-phenotype datasets, genetic ancestry and hereditary hearing loss, we find this framework outperforms several data-driven methods, particularly on low-data regimes. FREEFORM is available as open-source framework at GitHub: https://github.com/PennShenLab/FREEFORM.

Schlagwörter

genotype datafeature selectionfeature engineeringlarge language modelschain-of-thoughtensemble learninggenotype-phenotype predictiongenetic ancestryhereditary hearing losstabular data

Zitieren

APA

Li, Dawei, Duong-Tran, Duy, Wen, Z., Liu, Xiaoxi, Yang, Shu, Lee, Joseph, Chen, T., Hou, B., Shen, L., Tan, Zhen, & Baik, J. Y. (2025). Knowledge-Driven Feature Selection and Engineering for Genotype Data with Large Language Models. Carolina Digital Repository (University of North Carolina at Chapel Hill). https://doi.org/10.17615/fjd2-j454

Vancouver

Li, Dawei, Duong-Tran, Duy, Wen Z, Liu, et al. Knowledge-Driven Feature Selection and Engineering for Genotype Data with Large Language Models. Carolina Digital Repository (University of North Carolina at Chapel Hill). 2025. doi:10.17615/fjd2-j454

BibTeX
@article{Li2025,
  author = {Li and Dawei and Duong-Tran and Duy and Wen, Zixuan and Liu and Xiaoxi and Yang and Shu and Lee and Joseph and Chen, Tianlong and Hou, Bojian and Shen, Li and Tan and Zhen and Baik, Jae Young},
  title = {Knowledge-Driven Feature Selection and Engineering for Genotype Data with Large Language Models},
  journal = {Carolina Digital Repository (University of North Carolina at Chapel Hill)},
  year = {2025},
  doi = {10.17615/fjd2-j454}
}
Kostenlos starten →