文档介绍：Yeast Transcriptional Module Identification Based on
Line Manifold
Gangguo Li, Zhengzhi Wang
Institute of Automation, National University of Defense Technology, Changsha 410073, Hunan, China
ligangguo1982@
Abstract— A living cell can carry plex biological value<). Then it expands the core gene sets at a lower
functions dependent on its transcriptional regulatory modules. level of significance. ReMoDiscovery algorithm[5] takes
Therefore, identifying transcriptional regulatory modules is very similar stringent and relaxed two step procedures and infers
important for understanding cell function and its transcription TRMs from Chip-chip, motif and expression data. Module
mechanism. In this paper, a novel algorithm based on line Finding Algorithm (MOFA)[1] searches gene sets at a lower
manifold clusters and a linear mixture framework is presented level p-values with additional criteria for selecting genes
for transcriptional regulatory module identification. A regulated by a specific TF. Statistical-Algorithmic Method for
combined dissimilarity is constructed bining gene Bicluster Analysis (SAMBA) algorithm[6] transforms
expression and Chip-chip data with the linear mixture expression and Chip-chip data and generates a bipartite graph.
framework. Then bined dissimilarity is introduced into SAMBA requires discretization of inherently continuous