DomML, a novel ab initio protein domain prediction method based on domain size distribution

摘要：

BackgroundProtein domains represent a goldmine of information regarding tertiary structure prediction, protein classification, or functional inference. Thus predicting domains from data such as sequence and secondary structure is a crucial task. Indeed, its importance is such that a bi-annual competition is held to assess the performance of the different algorithms.State-of-the-art domain prediction algorithms rely on both template-based and ab initio approaches but the latter, which mostly use neural networks, are still unreliable as they exhibit low accuracy on multi-domain proteins.

作者: Yann Christinat

作者单位: Laboratory for Computational Biology and Bioinformatics, EPFL, 1015 Lausanne, Switzerland

会议类型: 国际会议

会议名称: The 7th Asia-Pacific Bioinformatics Conference(第七届亚太生物信息学大会)

会议地点: 北京

会议语种:英文

页码: 865

在线出版日期: 2009-01-01（万方平台首次上网日期，不代表论文的发表时间）