4.7 Article

YTLR: Extracting yeast transcription factor-gene associations from the literature using automated literature readers

期刊

出版社

ELSEVIER
DOI: 10.1016/j.csbj.2022.08.041

关键词

Transcriptional regulation; BioBERT; Natural language processing

资金

  1. National Cheng Kung University
  2. National Science and Technology Council of Taiwan [MOST 107-2218-E-390-009-MY3, MOST 110-2222-E-006-017, MOST 111-2221-E-006-231]

向作者/读者索取更多资源

Cells adapt to environmental stresses through transcription reprogramming, and the interactions between transcription factors (TF) and target genes play a key role in correct transcription control. YTLR is a pipeline tool that automates the extraction of TF-gene relations from literature, achieving high performance in the tasks.
Cells adapt to environmental stresses mainly via transcription reprogramming. Correct transcription control is mediated by the interactions between transcription factors (TF) and their target genes. These TFgene associations can be probed by chromatin immunoprecipitation techniques and knockout experiments, revealing TF binding (TFB) and regulatory (TFR) evidence, respectively. Nevertheless, most evidence is still fragmentary in the literature and requires tremendous human resources to curate. We developed the first pipeline called YTLR (Yeast Transcription-regulation Literature Reader) to automate TF-gene relation extraction from the literature. YTLR first identifies articles with TFB and TFR information. Then TF-gene binding pairs are extracted from the TFB articles, and TF-gene regulatory associations are recognized from the TFR papers. On gathered test sets, YTLR achieves an AUC value of 98.8% in identifying articles with TFB evidence and AUC = 83.4% in extracting the detailed TF-gene binding pairs. And similarly, YTLR also obtains an AUC value of 98.2% in identifying TFR articles and AUC = 80.4% in extracting the detailed TF-gene regulatory associations. Furthermore, YTLR outperforms previous methods in both tasks. To facilitate researchers in extracting TF-gene transcriptional relations from large-scale queried articles, an automated and easy-to-use software tool based on the YTLR pipeline is constructed. In summary, YTLR aims to provide easier literature pre-screening for curators and help researchers gather yeast TF-gene transcriptional relation conclusions from articles in a high-throughput fashion. The YTLR pipeline software tool can be downloaded at https://github.com/cobisLab/YTLR/. (c) 2022 The Author(s). Published by Elsevier B.V. on behalf of Research Network of Computational and Structural Biotechnology. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

作者

我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。

评论

主要评分

4.7
评分不足

次要评分

新颖性
-
重要性
-
科学严谨性
-
评价这篇论文

推荐

暂无数据
暂无数据