sREVEAL: Scalable extensions of REVEAL towards regulatory network inference

Vijender Chaitankar, P. Ghosh, M. Elasri, E. Perkins
{"title":"sREVEAL: Scalable extensions of REVEAL towards regulatory network inference","authors":"Vijender Chaitankar, P. Ghosh, M. Elasri, E. Perkins","doi":"10.1109/ISDA.2011.6121850","DOIUrl":null,"url":null,"abstract":"Most of the popular approaches towards gene regulatory networks inference e.g., Dynamic Bayesian Networks, Probabilistic Boolean Networks etc. are computationally complex and can only be used to infer small networks. While high-throughput experimental methods to monitor gene expression provide data for thousands of genes, these methods cannot fully utilize the entire spectrum of generated data. With the advent of information theoretic approaches in the last decade, the inference of larger regulatory networks from high throughput microarray data has become possible. Not all information theoretic approaches are scalable though; only methods that infer networks considering pair-wise interactions between genes such as, relevance networks, ARACNE and CLR to name a few, can be scaled upto genome-level inference. ARACNE and CLR attempt to improve the inference accuracy by pruning false edges, and do not bring in newer true edges. REVEAL is another information theoretic approach, which considers mutual information between multiple genes. As it goes beyond pair wise interactions, this approach was not scalable and could only infer small networks. In this paper, we propose two algorithms to improve the scalability of REVEAL by utilizing a transcription factor list (that can be predicted from the gene sequences) as prior knowledge and implementing time lags to further reduce the potential transcription factors that may regulate a gene. Our proposed S-REVEAL algorithms can infer larger networks with higher accuracy than the popular CLR algorithm.","PeriodicalId":433207,"journal":{"name":"2011 11th International Conference on Intelligent Systems Design and Applications","volume":"148 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2011-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"5","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2011 11th International Conference on Intelligent Systems Design and Applications","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ISDA.2011.6121850","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 5

Abstract

Most of the popular approaches towards gene regulatory networks inference e.g., Dynamic Bayesian Networks, Probabilistic Boolean Networks etc. are computationally complex and can only be used to infer small networks. While high-throughput experimental methods to monitor gene expression provide data for thousands of genes, these methods cannot fully utilize the entire spectrum of generated data. With the advent of information theoretic approaches in the last decade, the inference of larger regulatory networks from high throughput microarray data has become possible. Not all information theoretic approaches are scalable though; only methods that infer networks considering pair-wise interactions between genes such as, relevance networks, ARACNE and CLR to name a few, can be scaled upto genome-level inference. ARACNE and CLR attempt to improve the inference accuracy by pruning false edges, and do not bring in newer true edges. REVEAL is another information theoretic approach, which considers mutual information between multiple genes. As it goes beyond pair wise interactions, this approach was not scalable and could only infer small networks. In this paper, we propose two algorithms to improve the scalability of REVEAL by utilizing a transcription factor list (that can be predicted from the gene sequences) as prior knowledge and implementing time lags to further reduce the potential transcription factors that may regulate a gene. Our proposed S-REVEAL algorithms can infer larger networks with higher accuracy than the popular CLR algorithm.
sREVEAL: REVEAL对监管网络推理的可扩展扩展
大多数流行的基因调控网络推断方法,如动态贝叶斯网络,概率布尔网络等,计算复杂,只能用于推断小型网络。虽然监测基因表达的高通量实验方法提供了数千个基因的数据,但这些方法不能充分利用生成数据的整个频谱。随着近十年来信息理论方法的出现,从高通量微阵列数据推断更大的监管网络已经成为可能。并非所有的信息理论方法都是可扩展的;只有考虑到基因之间成对相互作用的网络推断方法,如相关网络、ARACNE和CLR等,才能扩大到基因组水平的推断。ARACNE和CLR试图通过修剪假边来提高推理精度,而不引入新的真边。揭示是另一种信息论方法,它考虑了多个基因之间的互信息。由于它超越了配对交互,因此这种方法不可扩展,只能推断小型网络。在本文中,我们提出了两种算法,通过利用转录因子列表(可以从基因序列中预测)作为先验知识和实施时间滞后来进一步减少可能调节基因的潜在转录因子来提高REVEAL的可扩展性。与流行的CLR算法相比,我们提出的S-REVEAL算法可以以更高的精度推断更大的网络。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
求助全文
约1分钟内获得全文 求助全文
来源期刊
自引率
0.00%
发文量
0
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术官方微信