We have developed a couple of algorithms and services, which are publically available.

SPEED - Finding Signatures of Signalling Pathways in Microarray data

SPEED is a signaling pathway annotation enrichment analysis tool with annotations based on evidences from pathway perturbation experiments. Thus, genes are annotated based on causal influences of pathway perturbations as opposed to pathway memberships alone. Identifying modulated pathways upstream of differentially expressed genes can facilitate the understanding of involved regulatory mechanisms. Currently only human genes and pathways are supported (developed together with Jignesh Parikh, Harvard & Boston University).

SPEED is described in the following publication:
Parikh, J, Klinger, B, Xia, Y, Marto, JA and Blüthgen, N. Discovering causal signaling pathways through gene expression patterns. Nucleic Acids Research, 38: W109-W117, 2010

TransFind - Finding transcription factors that regulate a set of co-expressed genes

TransFindIf you have a gene group (e.g. from gene expression studies) and you are interested whether there are any transcription factor binding sites enriched in the promoters of your gene group, this site may help you.

The prediction of which transcription factors regulate a group of target genes is based on affinities of the transcription factors to the putative promotors of the genes. The affinities have been precalculated based on the available positional frequency matrices for the transcription factors (developed together with Szymon Kielbasa, MPI MolGen).

Transfind is described in the following publication:
Kielbasa, SM, Klein, H., Roider, H., Vingron, M. and Blüthgen, N. TransFind - Predicting transcriptional regulators for gene sets. Nucleic Acids Research, 38: W275-W280, 2010

TargetFinder - Finding genes that show a similar expression across a large microarray panel

If you are looking for target genes of your favourite transcription factor this service could be helpful. Current methods are mainly based on sequence information alone. This classic way will produce a huge amount of false positives.. In order to get predictions that are much more likely to produce true positve results in your experimental work we provide a web tool that is based on large scale functional data as well as optionally based on binding statistics that will give you preselected candidates. Behind this functional data there are more than 1200 microarray experiments. So you can start your experimental work on genes that are more likey to be true targets because they are functionally very similar to your input set. That will save work and time (developed together with Ralf Mrowka, Charite, and Szymon Kielbasa, MPI MolGen).

Targetfinder is described in the following publication:
Kielbasa, SM, Blüthgen, N, Fahling, M, and Mrowka, R. a resource for systematic discovery of transcription factor target genes. Nucleic Acids Research, 38: W233-W238

Montecarlo-statistics on R x C matrices

Compares R x C table with randomized tables of same row and column totals. Developed to analyse bipartite network together with Nico Blüthgen, TU Darmstadt.

The algorithm has been used in the following publications:

Blüthgen, N., Menzel, F., Hovestadt, T., Fiala, B. and Blüthgen, N. Specialization, constraints, and conflicting interests in mutualistic networks. Curr Biol, 17 (4): 341-6, 2007

Blüthgen, N., Menzel, F. and Blüthgen, N. Measuring specialization in species interaction networks. BMC Ecol, 6: 9, 2006

Last modified:: 2013/03/16 10:50