Mining from protein–protein interactions

Abstract

Proteins are important cellular molecules, and interacting protein pairs provide biologically important information, such as functional relationships. We focus on the problem of predicting physically interacting protein pairs. This is an important problem in biology, which has been actively investigated in the field of data mining and knowledge discovery. Our particular focus is on data-mining-based methods, and the objective of this review is to introduce these methods for data mining researchers from technical viewpoints. We categorize those methods into three types: pairwise data-based, network-based, and integrative approaches, each approach being described in a different section. The first section is further divided into five types, such as supervised learning, algorithmic approaches, and unsupervised learning. The second section is mainly on link prediction, which can be further divided into two types, and two subsections that cover topics related with protein interaction networks are further added. The final section provides a wide variety of methods in integrative approaches. C 2012 Wiley Periodicals, Inc.

Research on the Link Prediction Model of Dynamic Multiplex Social Network Based on Improved Graph Representation Learning

“…As a complex network data analysis tool, link prediction can be used to process and mine various types of network information, such as assisting scientists in conducting biological protein structure analysis experiments to discover the interactions between different amino acids [1], helping merchants in product recommendation systems to recommend products and services to potential customers [2], assisting data engineers in data processing to retain hidden links and clean up false links [3]. At the same time, link prediction can be used as a criminal investigation tool for network mining and analysis of criminal suspects [4].…”

Link Prediction in Evolving Networks Based on Popularity of Nodes

“…For example, in protein-protein interaction experiments in cells, only strong relations between proteins could be detected by limited precision of equipments. It is prohibitive to measure every interaction between all pair proteins due to sharply increasing experimental costs with the size of proteins 7 , 8 , an appropriate approach is to evaluate the likelihood of potential relations and specifically test non-existing relations with the high likelihood. Also, in social contexts, two persons would build friendship in the near future with a high probability if they have many common friends or attributes, which could be utilized to uncover lost friends or predict future friends 9 – 11 .…”

Summarisation of weighted networks

“…Kimura, Saito, Nakano, and Motoda (2010)), hyperlinks between web pages (Lin, Yu, Han, and Liu (2010)), or interactions between proteins (e.g. Mamitsuka (2012)). In many of them, relationships have weights that are central to any use or analysis of networks: how frequently do two persons communicate, how much does web tra c flow from one page to another, or how strongly does one protein regulate the other one?…”