XJIPC OpenIR  > 多语种信息技术研究室
基于深度学习的蛋白质相互作用预测研究
王延斌
Thesis Advisor程力
2018-05-25
Degree Grantor中国科学院大学
Place of Conferral北京
Degree Name硕士
Degree Discipline计算机应用技术
Keyword蛋白质相互作用 深度学习 长短时记忆模型 位置特异性打分矩阵
Abstract

生命体的生理功能在很大程度上受到细胞中的蛋白质调控。细胞代谢的过程主要通过蛋白质复合物或者蛋白质与配体结合来完成的。研究表明蛋白质相互作用不仅主导着信号传导与识别、新陈代谢,还影响着复杂蛋白复合物的形成、细胞周期调整、癌症发生以及发展等重要过程。因此,蛋白质间的相互作用研究是一项非常重要的课题,这不仅有利于人类全面、深度的了解生命过程,而且有助于探索疾病的机制、开发生物制药以及精确寻找药物靶标。近些年,在实验室检测技术的高速发展下,蛋白质的序列数据进入了爆发式增长状态,如何有效使用这些蛋白质的序列信息,高效、准确的判断出蛋白质间是否会发生相互作用是目前亟待解决的难题。由于数据量大、实际的交互蛋白质网络又极其复杂导致传统实验的方法已经难以胜任今天的任务。此时,更高效、更灵活、更节省开支的机器学习方法给我们提供了另一个途径去解决这项前所未有的挑战。本文从蛋白质一级序列信息出发,开发了一套基于深度学习的蛋白质相互作用模型,该模型以深度长短时记忆神经网络为预测模型,并考虑了蛋白质序列信息内含的进化信息。该模型在两个蛋白质交互测试集上获得了令人满意的表现。为了证明深度学习的能力,我们比较了我们的方法和著名的支持向量机方法以及其他几种已知方法在相同数据集上的结果。结果表明,我们的方法在各项指标上都远超其他现有技术。

Other Abstract

The physiological functions of living organisms are largely regulated by proteins in cells. The process of cell metabolism is mainly accomplished by protein complexes or binding of proteins to ligands. It is shown that protein-protein interaction not only dominates signal transduction and recognition, metabolism, but also affects the formation of complex protein complexes, cell cycle regulation, cancer occurrence and development. Therefore, the study of protein-protein interaction is a very important topic, which is not only conducive to a comprehensive and in-depth understanding of human life processes, but also conducive to exploring the mechanism of disease, the development of biopharmaceuticals and the precise search for drug targets.In recent years, the rapid development of high-throughput technology has greatly promoted the explosive growth of protein sequence information. how to make good use of the rich protein sequence information and determine whether protein-protein interaction will occur efficiently and accurately is a difficult problem to be solved. Because of the large amount of data and the complexity of interactive protein networks, the traditional experimental methods are not suitable for today's tasks. At this point, a more efficient, flexible, and cost-effective approach to machine learning offers us another way to address this unprecedented challenge. Based on the first-order sequence information of proteins, a protein interaction model based on depth learning is developed. the model takes the deep long-term memory neural network as the prediction model and considers the evolution information contained in the protein sequence information.The model has been successfully demonstrated on two protein interaction test sets. To demonstrate the ability of in-depth learning, we compare the results of our method with the well-known support vector machine method and several other known methods on the same data set. The results show that our method is far superior to other existing techniques in every index.

Pages45
Document Type学位论文
Identifierhttp://ir.xjipc.cas.cn/handle/365002/5457
Collection多语种信息技术研究室
Recommended Citation
GB/T 7714
王延斌. 基于深度学习的蛋白质相互作用预测研究[D]. 北京. 中国科学院大学,2018.
Files in This Item:
File Name/Size DocType Version Access License
基于深度学习的蛋白质相互作用预测研究.p(2046KB)学位论文 开放获取CC BY-NC-SAApplication Full Text
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[王延斌]'s Articles
Baidu academic
Similar articles in Baidu academic
[王延斌]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[王延斌]'s Articles
Terms of Use
No data!
Social Bookmark/Share
All comments (0)
No comment.
 

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.