Exploiting inter-gene information for microarray data integration
Proceedings of the 2007 ACM symposium on Applied computing
A Hybrid DTW Based Method for Integration Analysis of Time Series Data
ICAIS '09 Proceedings of the 2009 International Conference on Adaptive and Intelligent Systems
Integrating heterogeneous microarray data sources using correlation signatures
DILS'05 Proceedings of the Second international conference on Data Integration in the Life Sciences
Clustering of multiple microarray experiments using information integration
ITBAM'11 Proceedings of the Second international conference on Information technology in bio- and medical informatics
| Hi-index | 0.00 |
In recent years, microarray gene expression profiles have become a common technique for inferring the relationship or regulation among different genes. While most of the previous work on microarray analysis focused on individual datasets, some global studies exploiting large numbers of microarrays have been presented recently. In this paper, we investigate how to integrate microarray data coming from different studies for the purpose of gene dependence analysis. In contrast to a meta-analysis approach, where results are combined on an interpretative level, we propose a method for direct integration analysis of gene relationships across different experiments and platforms. First, the algorithm utilizes a suitable metric in order to measure the relation between gene expression profiles. Then for each considered dataset a quadratic matrix that contains the interrelation values calculated between the expression profiles of each gene pair is constructed. Further a recursive aggregation algorithm is used in order to transform the set of constructed interrelation matrices into a single matrix, consisting of one overall inter-gene relation value per gene pair. At this stage a matrix of overall inter-gene relations obtained from previous data can be added and aggregated together with the currently constructed interrelation matrices. In this way, the previously generated integration results can, in fact, be updated with newly arriving ones studying the same phenomena. The obtained overall inter-gene relations can be considered as trade-off values agreed between the different experiments. These values express the gene correlation coefficients and therefore, may directly be analyzed in order to find the relationship among the genes.