I have done a cDNA microarray for cancer cells transduced with adeno5 vector holding my gene of interest and the control group was infected as well with adeno5 control ( empty vector) after 24h, over 14000 genes had fold changes after data analysis.
I know that my protein is signaling many things but did not expect a huge number of 14000 genes will be modified, How can I discarded the genes related to the adeno vector or the genes related to protein production and packaging, in order to minimize the genes I will analyze?
With the big data I have, I did different bioinformatics analysis e.g. Gsea and GO analysis, and still have a lot of data and pathways related to protein production or mRNA translation but I’m interested to look for genes related to cell death pathways and migration pathways (anticancer pathways in general).
From all the changed and enriched pathways, how can I know the ones related to my protein of interest and discard the ones that relate to cellular processes or considered as responses to the vector transduction.
14 000 DEGs is a huge number, really. Did you set filters (FDR, cut-off for fold change value) during statistical analysis of your data? In DAVID you can adjust the filter settings to generate restricted results.
does 1,5 fold change considered significant? and what FDR limit should i stay within?
Each fold change with correct p-value is significant. However, the higher FC cut-off you set, the shorter list you generate. It depends on you. If it comes about FDR - you shorten your list by setting lower values for FDR.
Hi,
FDR = 0,05 and Fc = +/- 2,00 are considered a standard, but it depends on you how stringent cut-off values you need. I used FDR 0,05 and Fc +/- 1,5, but I've seen also that people use FDR 0,1 for RNA-seq data.
Check if you can have access to Ingenuity Pathway Analysis. It´s expensive online tool, but maybe your University has access to it. It's very good to analyse biological functions, pathways and networks of your genes. When you start a new project you can go to advanced options and exclude the pathways/ biological functions/ deseases and other things you don't want to consider. In that way you can focus more on the things you are interested in.
Good luck :)
- Badges
- Science topic
- Similar topics
- Vectorization