Require review/comments/suggestions on 450k analysis college project

Question

[arpit.singh1203]

Hello everyone

I am a beginner level bioinformatician planning to work on "Analysing and interpreting the 450k methylation data for 70 odd samples from TCGA for a particular cancer". I intend to find DMRs when comparing tumor samples and normal samples. Later on I would like to co-relate the gene expression data with the methylation data.

Do you think this is a good research topic? Can you provide insight on what more can I do considering this is a six-months project? I will really appreciate your responses as I don't really have a 450k-savvy specialist/mentor anywhere around me.

Cheers

Answer 1

[Devon Ryan]

This is definitely an appropriate project, though how feasible it will be will depend a lot on what you background is and what sort of support and resources you have locally. Regarding what more you can do, the next obvious thing to look at would be functional enrichment of genes targeted by DMRs/expression changes. You might also want to look at clustering the samples, to see if there are any reasonable groupings (other than normal vs. cancer) that might provide some biological insite.

[arpit.singh1203]

Thank you very much for the comments and suggestions.

Answer 2

[Charles Warden]

I have some templates for identifying DMRs with 450k data here.

Also, COHCAP includes functions to integrate DNA methylation and gene expression data. However, you might want to have access to a Linux server if working with ~70 patient samples.

http://sourceforge.net/projects/cohcap/files/Protocol_Exchange_Example.zip

[arpit.singh1203]

Thanks a lot for the quick reply.