PhenotypeGene Relations (PGR) corpus
收藏资源简介:
PhenotypeGene Relations (PGR) corpus是由里斯本大学理学院的研究团队创建的一个银标数据集,专注于人类表型与基因关系的标注。该数据集包含1712篇摘要,涉及5676个人类表型标注和13835个基因标注,以及4283个关系标注。创建过程中使用了命名实体识别工具,并通过八位编者部分评估,达到87.01%的精确度。数据集主要用于解决生物医学文献中自动提取表型-基因关系的问题,支持深度学习工具的开发和评估。
The PhenotypeGene Relations (PGR) corpus is a silver-standard dataset created by a research team from the Faculty of Science, University of Lisbon, focusing on the annotation of human phenotype-gene relationships. This corpus contains 1,712 abstracts, involving 5,676 human phenotype annotations, 13,835 gene annotations, and 4,283 relationship annotations. During its creation, named entity recognition tools were utilized, and partial evaluation was conducted by eight annotators, achieving an accuracy of 87.01%. This dataset is primarily used to address the task of automatically extracting phenotype-gene relationships from biomedical literature, supporting the development and evaluation of deep learning tools.




