MetaPred2CS：一种基于序列的原核二组分系统蛋白蛋白质相互作用的元预测器。

MetaPred2CS: a sequence-based meta-predictor for protein-protein interactions of prokaryotic two-component system proteins.

机构信息

Institute of Biological, Environmental and Rural Sciences, Aberystwyth University, Aberystwyth SY23 3EB, UK.

出版信息

Bioinformatics. 2016 Nov 1;32(21):3339-3341. doi: 10.1093/bioinformatics/btw403. Epub 2016 Jul 4.

DOI:10.1093/bioinformatics/btw403

PMID:27378293

Abstract

MOTIVATION

Two-component systems (TCS) are the main signalling pathways of prokaryotes, and control a wide range of biological phenomena. Their functioning depends on interactions between TCS proteins, the specificity of which is poorly understood.

RESULTS

The MetaPred2CS web-server interfaces a sequence-based meta-predictor specifically designed to predict pairing of the histidine kinase and response-regulator proteins forming TCSs. MetaPred2CS integrates six sequence-based methods using a support vector machine classifier and has been intensively tested under different benchmarking conditions: (i) species specific gene sets; (ii) neighbouring versus orphan pairs; and (iii) k-fold cross validation on experimentally validated datasets.

AVAILABILITY AND IMPLEMENTATION

Web server at: http://metapred2cs.ibers.aber.ac.uk/, Source code: https://github.com/martinjvickers/MetaPred2CS or implemented as Virtual Machine at: http://metapred2cs.ibers.aber.ac.uk/download CONTACT: naf4@aber.ac.ukSupplementary information: Supplementary data are available at Bioinformatics online.

摘要

动机

双组分系统（TCS）是原核生物的主要信号通路，控制着广泛的生物现象。它们的功能取决于 TCS 蛋白之间的相互作用，而这种相互作用的特异性还不太清楚。

结果

MetaPred2CS 网络服务器接口是一个基于序列的元预测器，专门用于预测形成 TCS 的组氨酸激酶和反应调节剂蛋白的配对。MetaPred2CS 集成了六种基于序列的方法，使用支持向量机分类器，并在不同的基准测试条件下进行了深入测试：（i）特定于物种的基因集；（ii）相邻对与孤儿对；以及（iii）在实验验证数据集上进行 k 折交叉验证。