Limits...
An integrative clustering and modeling algorithm for dynamical gene expression data.

Sivriver J, Habib N, Friedman N - Bioinformatics (2011)

Bottom Line: Moreover, our approach provides an easy way to compare between responses to different stimuli at the dynamical level.We use our approach to analyze the dynamical transcriptional responses to inflammation and anti-viral stimuli in mice primary dendritic cells, and extract a concise representation of the different dynamical response types.We analyze the similarities and differences between the two stimuli and identify potential regulators of this complex transcriptional response.

View Article: PubMed Central - PubMed

Affiliation: School of Computer Science and Engineering, The Hebrew University of Jerusalem, Jerusalem 91904, Israel.

ABSTRACT

Motivation: The precise dynamics of gene expression is often crucial for proper response to stimuli. Time-course gene-expression profiles can provide insights about the dynamics of many cellular responses, but are often noisy and measured at arbitrary intervals, posing a major analysis challenge.

Results: We developed an algorithm that interleaves clustering time-course gene-expression data with estimation of dynamic models of their response by biologically meaningful parameters. In combining these two tasks we overcome obstacles posed in each one. Moreover, our approach provides an easy way to compare between responses to different stimuli at the dynamical level. We use our approach to analyze the dynamical transcriptional responses to inflammation and anti-viral stimuli in mice primary dendritic cells, and extract a concise representation of the different dynamical response types. We analyze the similarities and differences between the two stimuli and identify potential regulators of this complex transcriptional response.

Availability: The code to our method is freely available http://www.compbio.cs.huji.ac.il/DynaMiteC.

Contact: nir@cs.huji.ac.il.

Show MeSH
Comparing two stimuli. (A) Distribution of clusters reassignment in the polyIC response (PIC clusters) when compared to the cluster assignment in the LPS response. For each LPS cluster (X-axis), we see the distribution of genes (Y-axis) to polyIC clusters (colored bars). (B) Examples of the differences in dynamic response of a gene under LPS (straight line) and polyIC (dashed line) stimuli, and their assignment to clusters (color of the line). TNF (left) has lower response in polyIC (Cluster 2) compared to LPS (Cluster 1). FABP3 (middle) has a delayed response in polyIC (Cluster 3) compared to LPS (Cluster2). CXCL10 (right) has a very similar response to both stimuli (Cluster 1).
© Copyright Policy - creative-commons
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC3117368&req=5

Figure 4: Comparing two stimuli. (A) Distribution of clusters reassignment in the polyIC response (PIC clusters) when compared to the cluster assignment in the LPS response. For each LPS cluster (X-axis), we see the distribution of genes (Y-axis) to polyIC clusters (colored bars). (B) Examples of the differences in dynamic response of a gene under LPS (straight line) and polyIC (dashed line) stimuli, and their assignment to clusters (color of the line). TNF (left) has lower response in polyIC (Cluster 2) compared to LPS (Cluster 1). FABP3 (middle) has a delayed response in polyIC (Cluster 3) compared to LPS (Cluster2). CXCL10 (right) has a very similar response to both stimuli (Cluster 1).

Mentions: An advantage of Impulse clustering is that it provides an easy way to compare between different transcriptional responses. We can apply our method simultaneously to different time series data, and test for each gene, not only in which of the responses is it involved, but also compare its dynamical expression profile across conditions. Note that since we model the expression of each gene by a continuous function, we can compare experiments measuring different time points. We compared between the response to LPS (inflammation) and polyIC (viral) stimuli using our method. In general, most of the genes respond to both stimuli, while only 148 genes are specifically anti-viral, and 110 are LPS specific. Among the shared genes, we can easily find genes that change their dynamical response between stimuli, by looking at genes classified to a certain cluster in LPS and a different cluster in polyIC (Fig. 4a). In general, 65% of the genes change their dynamical response between the two stimuli. We focused on sets of genes that change their dynamical response in a similar manner and characterized them by enrichments of gene annotations (hypergeometeric P-value with Bonferroni correction). Interestingly, the 286 genes that are classified as primary response genes (Clusters 1 or 2) in LPS and secondary response in polyIC (Clusters 3 or 4) are enriched for immune response genes (P<2×10−5), signaling cascade genes (P<6×10−4), and positive regulation of apoptosis (P<9×10−3). While the genes that maintain their cluster assignment in Cluster 1 are enriched for immune response (P<8×10−5) and anti-viral response (P<9×10−3), and these that remain in Cluster 2, are enriched for transcription regulation (P<6×10−4).Fig. 4.


An integrative clustering and modeling algorithm for dynamical gene expression data.

Sivriver J, Habib N, Friedman N - Bioinformatics (2011)

Comparing two stimuli. (A) Distribution of clusters reassignment in the polyIC response (PIC clusters) when compared to the cluster assignment in the LPS response. For each LPS cluster (X-axis), we see the distribution of genes (Y-axis) to polyIC clusters (colored bars). (B) Examples of the differences in dynamic response of a gene under LPS (straight line) and polyIC (dashed line) stimuli, and their assignment to clusters (color of the line). TNF (left) has lower response in polyIC (Cluster 2) compared to LPS (Cluster 1). FABP3 (middle) has a delayed response in polyIC (Cluster 3) compared to LPS (Cluster2). CXCL10 (right) has a very similar response to both stimuli (Cluster 1).
© Copyright Policy - creative-commons
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC3117368&req=5

Figure 4: Comparing two stimuli. (A) Distribution of clusters reassignment in the polyIC response (PIC clusters) when compared to the cluster assignment in the LPS response. For each LPS cluster (X-axis), we see the distribution of genes (Y-axis) to polyIC clusters (colored bars). (B) Examples of the differences in dynamic response of a gene under LPS (straight line) and polyIC (dashed line) stimuli, and their assignment to clusters (color of the line). TNF (left) has lower response in polyIC (Cluster 2) compared to LPS (Cluster 1). FABP3 (middle) has a delayed response in polyIC (Cluster 3) compared to LPS (Cluster2). CXCL10 (right) has a very similar response to both stimuli (Cluster 1).
Mentions: An advantage of Impulse clustering is that it provides an easy way to compare between different transcriptional responses. We can apply our method simultaneously to different time series data, and test for each gene, not only in which of the responses is it involved, but also compare its dynamical expression profile across conditions. Note that since we model the expression of each gene by a continuous function, we can compare experiments measuring different time points. We compared between the response to LPS (inflammation) and polyIC (viral) stimuli using our method. In general, most of the genes respond to both stimuli, while only 148 genes are specifically anti-viral, and 110 are LPS specific. Among the shared genes, we can easily find genes that change their dynamical response between stimuli, by looking at genes classified to a certain cluster in LPS and a different cluster in polyIC (Fig. 4a). In general, 65% of the genes change their dynamical response between the two stimuli. We focused on sets of genes that change their dynamical response in a similar manner and characterized them by enrichments of gene annotations (hypergeometeric P-value with Bonferroni correction). Interestingly, the 286 genes that are classified as primary response genes (Clusters 1 or 2) in LPS and secondary response in polyIC (Clusters 3 or 4) are enriched for immune response genes (P<2×10−5), signaling cascade genes (P<6×10−4), and positive regulation of apoptosis (P<9×10−3). While the genes that maintain their cluster assignment in Cluster 1 are enriched for immune response (P<8×10−5) and anti-viral response (P<9×10−3), and these that remain in Cluster 2, are enriched for transcription regulation (P<6×10−4).Fig. 4.

Bottom Line: Moreover, our approach provides an easy way to compare between responses to different stimuli at the dynamical level.We use our approach to analyze the dynamical transcriptional responses to inflammation and anti-viral stimuli in mice primary dendritic cells, and extract a concise representation of the different dynamical response types.We analyze the similarities and differences between the two stimuli and identify potential regulators of this complex transcriptional response.

View Article: PubMed Central - PubMed

Affiliation: School of Computer Science and Engineering, The Hebrew University of Jerusalem, Jerusalem 91904, Israel.

ABSTRACT

Motivation: The precise dynamics of gene expression is often crucial for proper response to stimuli. Time-course gene-expression profiles can provide insights about the dynamics of many cellular responses, but are often noisy and measured at arbitrary intervals, posing a major analysis challenge.

Results: We developed an algorithm that interleaves clustering time-course gene-expression data with estimation of dynamic models of their response by biologically meaningful parameters. In combining these two tasks we overcome obstacles posed in each one. Moreover, our approach provides an easy way to compare between responses to different stimuli at the dynamical level. We use our approach to analyze the dynamical transcriptional responses to inflammation and anti-viral stimuli in mice primary dendritic cells, and extract a concise representation of the different dynamical response types. We analyze the similarities and differences between the two stimuli and identify potential regulators of this complex transcriptional response.

Availability: The code to our method is freely available http://www.compbio.cs.huji.ac.il/DynaMiteC.

Contact: nir@cs.huji.ac.il.

Show MeSH