Finding Time-Lagged 3D Clusters

Existing 3D clustering algorithms on gene ? sample ? time expression data do not consider the time lags between correlated gene expression patterns. Besides, they either ignore the correlation on time subseries, or disregard the continuity of the time series, or only validate pure shifting or pure scaling coherent patterns instead of the general shiftingand-scaling patterns. In this paper, we propose a novel 3D cluster model, S2 D3 Cluster, to address these problems, where S2 reflects the shifting-and-scaling correlation and D3 the 3Dimensional gene?sample?time data. Within the S2 D3 Cluster model, expression levels of genes are shifting-and-scaling coherent in both sample subspace and time subseries with arbitrary time lags. We develop a 3D clustering algorithm, LagMiner, for identifying interesting S2 D3 Clusters that satisfy the constraints of regulation (), coherence ( ), minimum gene number (MinG), minimum sample subspace size (MinS) and minimum time periods length (MinT). Experim...

• Type: Conference
• Year: 2009
• Where: ICDE
• Authors: Xin Xu, Ying Lu, Kian-Lee Tan, Anthony K. H. Tung
• Keywords: Arbitrary Time Lags, Database, Gene Expression Data, Gene Expression Patterns, ICDE 2009, Minimum Time Periods, Time Expression Data