In higher plants, gene expression of the 20.000 to > 50.000 genes is regulated by 2.000 to 4.000 transcription factors (TFs). Although the functions of several TFs have been discovered, the target genes of TFs and hence the gene regulatory networks they control is unknown for the vast majority of them. Currently, more and more ChIP-Seq data are collected from Arabidopsis thaliana and very likely crops in the near future as well. Other experimental data are collected as well, from EMSA experiments, binding site selection assays, transactivation assays, inducible expression of TFs followed by microarray hybridization or RNAseq to identify genes affected by the TFs.

Currently, it is very time-consuming to collect and combine the various data needed to unravel the GRNs of transcription factors. Experimentalists are collecting information from diverse sources and then try to combine the extracted information to establish the GRN.

nd

There are various tools available allow for coexpression analysis, motif ifentification (e.g. Patmatch), the extraction of promoters from different plant specie (e.g. phytozome), phylogenetic footprinting, cis element databases, etc. However, it is extremely time-consuming to transfer data from A to B and extract the information that is needed to establish a GRN.

• CORNET. https://cornet.psb.ugent.be/main/precalc
  
• ATTED-II . http://atted.jp/top_search.shtml#CoExSearch
  
• CressExpress. http://cressexpress.org/
  
• CSB.DB http://csbdb.mpimp-golm.mpg.de/
  
• Patmatch in TAIR
  
• AGRIS, TransFac, PLACE for the identification of known cis elements in promoters of interest (e.g. of coexpressed genes).
  
• MEME for the identification of motifs overrepresented in a set of promoters.

Drawbacks: The main drawback is that it is extremely time-consuming to combine the different data into one single scheme that displays the GRN.

We need user-friendly tools that enables the efficient reconstruction of GRNs from experimental and computational data. For example, I want to know the most likely direct target genes of the given TF, starting out from genes that (at their expression level) respond to a change in the activity of a given TF.

Protocol:

1. Perform expression analysis (Affy, Agilent, RNAseq) after activation of a TF
2. Identify responding genes
3. Extract their promoter sequences
4. Identify known cis elements in the promoters of these genes
5. Identify novel motifs in the promoters of these genes
6. Check whether binding sites of the TF under analysis is present in the promoters of the responding genes; at which position, in which number? On which strand of the DNA? Etc.
7. Compare the promoters of these genes with promoters of orthologous genes from other plant species (crops!)
8. Use phylogenetic information to cross-check for cis elements / motifs.
9. Include expression data from other experiments in Arabidopsis and other species
10. Suggest a GRN in a good schematic presentation that highlights various aspects of the interaction of the TF with its downstream target genes (e.g. cis elements in the promoters to which it might bind; expression induction or repression curve of the target gene);

LF: Possible person to invite: Chris J Needham C.Needham@leeds.ac.uk (see http://www.biomedcentral.com/1752-0509/3/85)