Limits...
Response projected clustering for direct association with physiological and clinical response data.

Yi SG, Park T, Lee JK - BMC Bioinformatics (2008)

Bottom Line: We have shown that RPC can effectively discover gene networks with different degrees of association with clinical metadata.Performed on each gene's response projected vector based on its degree of association with the response data, RPC effectively summarizes individual genes' association with metadata as well as their own expression patterns.Thus, RPC greatly enhances the utility of clustering analysis on investigating high-dimensional microarray gene expression data with quantitative metadata.

View Article: PubMed Central - HTML - PubMed

Affiliation: Department of Statistics, Seoul National University, Silim-dong, Kwanak-gu, Seoul, 151-747, Korea. skon@bibs.snu.ac.kr

ABSTRACT

Background: Microarray gene expression data are often analyzed together with corresponding physiological response and clinical metadata of biological subjects, e.g. patients' residual tumor sizes after chemotherapy or glucose levels at various stages of diabetic patients. Current clustering analysis cannot directly incorporate such quantitative metadata into the clustering heatmap of gene expression. It will be quite useful if these clinical response data can be effectively summarized in the high-dimensional clustering display so that important groups of genes can be intuitively discovered with different degrees of relevance to target disease phenotypes.

Results: We introduced a novel clustering analysis approach, response projected clustering (RPC), which uses a high-dimensional geometrical projection of response data to the gene expression space. The projected response vector, which becomes the origin in the projected space, is then clustered together with the projected gene vectors based on their different degrees of association with the response vector. A bootstrap-counting based RPC analysis is also performed to evaluate statistical tightness of identified gene clusters. Our RPC analysis was applied to the in vitro growth-inhibition and microarray profiling data on the NCI-60 cancer cell lines and the microarray gene expression study of macrophage differentiation in atherogenesis. These RPC applications enabled us to identify many known and novel gene factors and their potential pathway associations which are highly relevant to the drug's chemosensitivity activities and atherogenesis.

Conclusion: We have shown that RPC can effectively discover gene networks with different degrees of association with clinical metadata. Performed on each gene's response projected vector based on its degree of association with the response data, RPC effectively summarizes individual genes' association with metadata as well as their own expression patterns. Thus, RPC greatly enhances the utility of clustering analysis on investigating high-dimensional microarray gene expression data with quantitative metadata.

Show MeSH

Related in: MedlinePlus

Co-clustering counting. The q-percentile node for co-clustering counting.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC2275250&req=5

Figure 8: Co-clustering counting. The q-percentile node for co-clustering counting.

Mentions: Due to the nature of its heuristic allocation algorithms, clustering analysis can often provide different groups of clustered genes with slightly different input data or even with different orders of genes. Statistical confidence evaluation on clustered gene groups has thus been suggested using resampling techniques such as bootstrap [28,29]. We also use a bootstrapping technique to assess the stability of our RPC clustering results among RPC selected, say, s genes. We obtain B bootstrapped samples of size n {z1b, ..., znb}, b = 1,...,B from the original n subjects (column vectors) {z1*, ..., zn*} with replacement where zj* = {x1i*, ..., xpi*} is the s-dimensional vector of the j-th subject. The consistency of sub-clusters of the s genes can be examined from these bootstrapped samples. For example, the probability that two genes belong to a common subcluster can be assessed by counting the frequencies of their co-clustering occurrences at a particular node, e.g. 75-percentile node of each cluster dendrogram (Fig. 8).


Response projected clustering for direct association with physiological and clinical response data.

Yi SG, Park T, Lee JK - BMC Bioinformatics (2008)

Co-clustering counting. The q-percentile node for co-clustering counting.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC2275250&req=5

Figure 8: Co-clustering counting. The q-percentile node for co-clustering counting.
Mentions: Due to the nature of its heuristic allocation algorithms, clustering analysis can often provide different groups of clustered genes with slightly different input data or even with different orders of genes. Statistical confidence evaluation on clustered gene groups has thus been suggested using resampling techniques such as bootstrap [28,29]. We also use a bootstrapping technique to assess the stability of our RPC clustering results among RPC selected, say, s genes. We obtain B bootstrapped samples of size n {z1b, ..., znb}, b = 1,...,B from the original n subjects (column vectors) {z1*, ..., zn*} with replacement where zj* = {x1i*, ..., xpi*} is the s-dimensional vector of the j-th subject. The consistency of sub-clusters of the s genes can be examined from these bootstrapped samples. For example, the probability that two genes belong to a common subcluster can be assessed by counting the frequencies of their co-clustering occurrences at a particular node, e.g. 75-percentile node of each cluster dendrogram (Fig. 8).

Bottom Line: We have shown that RPC can effectively discover gene networks with different degrees of association with clinical metadata.Performed on each gene's response projected vector based on its degree of association with the response data, RPC effectively summarizes individual genes' association with metadata as well as their own expression patterns.Thus, RPC greatly enhances the utility of clustering analysis on investigating high-dimensional microarray gene expression data with quantitative metadata.

View Article: PubMed Central - HTML - PubMed

Affiliation: Department of Statistics, Seoul National University, Silim-dong, Kwanak-gu, Seoul, 151-747, Korea. skon@bibs.snu.ac.kr

ABSTRACT

Background: Microarray gene expression data are often analyzed together with corresponding physiological response and clinical metadata of biological subjects, e.g. patients' residual tumor sizes after chemotherapy or glucose levels at various stages of diabetic patients. Current clustering analysis cannot directly incorporate such quantitative metadata into the clustering heatmap of gene expression. It will be quite useful if these clinical response data can be effectively summarized in the high-dimensional clustering display so that important groups of genes can be intuitively discovered with different degrees of relevance to target disease phenotypes.

Results: We introduced a novel clustering analysis approach, response projected clustering (RPC), which uses a high-dimensional geometrical projection of response data to the gene expression space. The projected response vector, which becomes the origin in the projected space, is then clustered together with the projected gene vectors based on their different degrees of association with the response vector. A bootstrap-counting based RPC analysis is also performed to evaluate statistical tightness of identified gene clusters. Our RPC analysis was applied to the in vitro growth-inhibition and microarray profiling data on the NCI-60 cancer cell lines and the microarray gene expression study of macrophage differentiation in atherogenesis. These RPC applications enabled us to identify many known and novel gene factors and their potential pathway associations which are highly relevant to the drug's chemosensitivity activities and atherogenesis.

Conclusion: We have shown that RPC can effectively discover gene networks with different degrees of association with clinical metadata. Performed on each gene's response projected vector based on its degree of association with the response data, RPC effectively summarizes individual genes' association with metadata as well as their own expression patterns. Thus, RPC greatly enhances the utility of clustering analysis on investigating high-dimensional microarray gene expression data with quantitative metadata.

Show MeSH
Related in: MedlinePlus