Entering edit mode
7.7 years ago
rvrob3000
•
0
Hi everyone,
I have been looking for a tool in order to build a gene regulatory network. Long story short, I have a set of Arabidopsis gene identifiers and I would want to upload these gene identifiers into a program that would be able to create a gene regulatory network for me. Unfortunately, all the programs I have found require some sort of expression data, which I do not have. So does anyone know of a gene regulatory network tool that can create regulatory networks from gene identifiers or at least uniprot accession numbers only without needing to have expression data?
Thank You,
If it was as easy as getting a list of gene IDs and nothing else .... we would have solved the worlds health problems 100 years ago.
Do you have anything at all that links those genes? have you randomly selected them? where did the list come from?
Hi Kenneth,
No I don't have anything that links the genes. The list would come from a large proteomics data set. I know they are transcription factors from the GO annotations and TAIR descriptions.
A "large proteomics dataset" could mean a lot of things. Input the gene identifiers into StringDB and you will see the kind of network you get. My guess is that if the gene list is not filtered to included only those IDs with some biological connection that you are testing (e.g. co-expression or tissue specificity) then your result will just be a load of nodes with very few edges.
An alternative is i-regulon which may give similar results.
However, you may get lucky! So give them a try.
Searching in ppi database is a good place to start with, since the questioner doesn't have any information about how two genes should be connected. This will work if the questioner's gene regulatory network is equal to ppi network.
Maybe you could try to map some of your genes to pathways like in KEGG and then you will be able to plot a regulatory network.
That would be good.... but how would you select "some" of the genes? IF random....you'd have to be pretty lucky to find anything. "Some" of the genes are usually selected by co-expression .... the filtering/selection process is the key to getting the network. The tool that builds the network is just a tool..and it follows the same rules as all other tools...."Rubbish In, Rubbish Out".
Hi solo7773,
I tried that. A lot of them don't have KO's assigned to them. For example if I were to search for "AT1G30330", it doesn't have a K number and there are no pathways assigned to it.
http://www.genome.jp/dbget-bin/www_bget?ath:AT1G30330
That indeed happens as the database only stores the knowledge we've learned. Another choice, maybe you can try to search a bioinformatic tool that can generate/predict a regulatory network (or regulatory interactions) based on gene list. After that, you will be able to use any tools like cytoscape to make a network visualisation.