文档介绍：work-based Approach for Protein Functions
Prediction Using Locally Linear Embedding
Haifeng Zhao Dengdi Sun Rifeng Wang Bin Luo
Key Lab of puting & Signal Processing of Ministry of Education
School puter Science and Technology, Anhui University, Hefei, 230039, P. R. China
E-mail: senith@ sundengdi@
Abstract—Inferring protein functions from different data sources expression, but can predict that two proteins share a function
is a challenging task in the post-genomic era, as a large number even when they have no sequence similarity. For example,
of crude protein structures from structural genomics project are gene fusion methods [1], ic profiles [2], Genomic
now solved without their biochemical functions characterized. context methods [3]. However, discrepancies of prediction may
Recently, many different methods have been used to predict arise due to the corruptions of gene expression data or protein
protein functions including those based on Protein-Protein sequences. Moreover, occasionally, there are corruptions or
Interaction (PPI), structure, sequence relationship, gene
expression data, etc. Among these approaches, methods based on bad probes in microarrays and protein sequence.
protein interaction data are very promising. In this paper, we Protein-Protein Interaction (PPI) plays a key role in many
studied work-based method using local