Biostatistics

Knowledge-based matrix factorization temporally resolves the cellular responses to IL-6 stimulation

External stimulations of cells by hormones, cytokines or growth factors activate signal transduction pathways that subsequently induce a re-arrangement of cellular gene expression. The representation and analysis of changes in gene response is complicated, as they essentially consist of multi-layered temporal responses. In such situation, matrix factorization techniques provide efficient tools for the detailed temporal analysis. Related methods applied in bioinformatics so far did not take prior knowledge into account. In other fields, factorization techniques incorporating data properties like second-order spatial and temporal structures have shown a robust performance. However, large-scale biological data rarely imply a natural order that allows the definition of an autocorrelation function.

We therefore develop the concept of graph–decorrelation. We encode prior knowledge like transcriptional regulation, protein interactions or metabolic pathways as a weighted directed graph. By linking features along this underlying graph, we introduce a partial ordering of the samples to define an autocorrelation function. Using this framework as constraint to the matrix factorization task allows us to set up the fast and robust graph–decorrelation algorithm (GraDe). To analyze alterations in the gene response in IL-6 stimulated primary mouse hepatocytes by GraDe, a time-course microarray experiment was performed. Extracted time–resolved gene expression profiles show that IL-6 activates genes involved in cell cycle progression and cell division. On the contrary, genes linked to metabolic and apoptotic processes are down-regulated indicating that IL-6 mediated priming renders hepatocytes more responsive towards cell proliferation and reduces expenditures for the energy metabolism.

 

References

  • Kowarsch A, Blöchl F, Bohl S, Saile M, Gretz N, Klingmüller U, Theis FJ: Knowledge-based matrix factorization temporally resolves the cellular responses to IL-6 stimulation. BMC Bioinformatics, 11:585, 2010 [PubMedDOI]
  • Blöchl F, Kowarsch A, Theis FJ: Second-order source separation based on prior knowledge realized in a graph modelIn Proc. LCA/ICA 2010, Lecture Notes of Computer Science, Springer, 2010
  • Raia V, Schilling M, Böhm M, Hahn B, Kowarsch A, Raue A, Sticht C, Bohl S, Saile M, Möller P, Gretz N, Timmer J, Theis FJ, Lehmann WD, Lichter P, Klingmüller U: Dynamic Mathematical Modeling of IL13-induced Signaling in Hodgkin and Primary Mediastinal B-cell Lymphoma Allows Prediction of Therapeutic Targets. Cancer Research in press [PubMed]

Downloads

Graph decorrelation (Grade) algorithm:

grade.m  for file input, e.g genes.txt and network.txt

grade_core.m  for matrices, e.g. stored in bifan.mat or funnel.mat

 

Supplementary Material

IL-6 stimulated primary mouse hepatocytes :

genes.txt  log2 gene expression values

network.txt  underlying gene regulatory network

source.xls  g-decorrelated sources

Toy examples :

bifan.mat  Matlab files contains expression and network information for the bifan toy example

funnel.mat  Matlab files contains expression and network information for the funnel toy examples

We use cookies to improve your experience on our Website. We need cookies to continuously improve the services, to enable certain features and when embedding services or content of third parties, such as video player. By using our website, you agree to the use of cookies. We use different types of cookies. You can personalize your cookie settings here:

Show detail settings
Please find more information in our privacy statement.

There you may also change your settings later.