Wgcna-Trait File For Microarray Data

Question

[pixie@bioinfo]

Hi, I am working on plant microarray data in drought condition. I am using WGCNA (R package) for building co-expression modules. In the tutorial, they have mentioned about clinical trait data. I wanted to know what kind of data I can use in my case ? fold change ?..any suggestions will be very much appreciated.

[hossein.banabazi]

Dear Sanchari,

You can find differently expressed genes from your data by limma package and then cluster them by WGCNA. There is a classic example on Yeast Cell Cycle data. I wrote R codes and have required data. Let me know if you are interested.

Best Regards,

Mohammad

Mohammad Hossein Banabazi PhD candidate in Animal Genetics & Breeding (Univ. of Tehran), & Member of Scientific Board

Department of Biotechnology, Animal Science Research Institute of IRAN (ASRI), The First Dehghan Villa, Shahid Beheshti St., P.O.Box: 1483, Post Code: 3146618361, Karaj, IRAN. Tel (ASRI): 98-26-34430010 (ext. 467) Tel (Home): 98-26-34470775 Mobile: [phone] Fax: [phone] Alternative E-mail: [e-mail], [e-mail], [e-mail] and [e-mail] ID: ooVoo & Skype: banabazi, Yahoo Messenger: mh_banabazi, Google Talk: hossein.banabazi Webpage: http://www.asri.ir/Portal/File/ShowFile.aspx?ID=e72f8b94-bd99-460b-b119-b08116c3dacb CV link: http://www.nrteam.ir/C.V%20NRTeam%20members/C.V%20MOHAMMAD%20HOSSEIN%20BANABAZI.pdf

[rajasekargutha]

please sir, could you provide r code, i am troubling to match trait data with its expression data

Answer 1

[Eric Normandeau]

WGCNA is used to create clusters of gene (modules) that are co-expressed among your samples and then to correlate the modules with a trait/condition to find groups of genes that seem to vary in expression with changes in your trait/condition. Finding the co-expression of your modules with fold-change is a strange idea.

I am not sure what this would tell you. You would be trying to find gene modules that vary in expression. I suppose you probably already have a statistical way of finding differently expressed genes from your data.

If I remember well, I think you can use discreet traits, so you could try to correlate with drought condition (drought, no-drought). You can test multiple traits, so even better would be if you had different traits to test. For example: percent of leaves area that is dried (necrosed), millimetres of rain per day, number of days without water, etc.

[pixie@bioinfo]

Thanks for the insight..the tutorials don't say much about the trait file. I'll try using the discreet trait that you mentioned about..

[bioLife]

Hi! So in the case of drought/non-drought, one can create modules separately and afterwards compare the modules in the two conditions? eg.find modules which have same expression patterns?
Or should create modules using both conditions together?