If I have a ran-seq data of disease A and also rna-seq for disease B on the same platform and If i extracted DEGs from both diseases and find common DEGs; is this analysis ok?

Question

[Md. Al - Mustanjid]

If I have a ran-seq data of disease A and also rna-seq for disease B on the same platform and If i extracted DEGs from both diseases and find common DEGs; is this analysis ok?

Answer 1

[Kevin Blighe]

Yes; however, the way that you report on the results (the overlapping / common genes) is important. You could 'formalise' the overlap by doing a meta analysis.

Also, if all samples were profiled together in the same experiment, it would be better to process and normalise these together, and also perform the differential expression comparisons.

Kevin

[Md. Al - Mustanjid]

Thank you a lot Kevin. It was quite impossible to conduct my research work without your help. Your suggestions make my study path more clear. I am moving forward to ran-seq analysis, as I am a beginner I have a hard a lot about meta-analysis. But I don't have proper knowledge on this. could you please suggest me any source or something like that. It will be more helpful for my learning process. Thanks again.

[Kevin Blighe]

Sure, but, are these diseases completely different or are they sub-types of the same disease? It is not a problem to simply compare the 2 lists of the genes that are statistically significant. What do you ultimately hope to infer from this particular part of the analysis?

For a simple meta-analysis, I found this program easy to use: https://www.bioconductor.org/packages/release/bioc/html/GenRank.html

You could also compare the fold-changes of all genes in a scatter plot, i.e., log2FC Disease A (x-axis) versus log2FC Disease B (y-axis), and derive a Pearson correlation co-efficient and p-value for this.

[Md. Al - Mustanjid]

Thanks Kevin. Disease A is a subtype of disease B. I want to detect molecular insights and signatures between these diseases from DEGs shared in both diseases from Systems biology perspective. I will check out the package.

[Kevin Blighe]

Cool. You should be doing pathway analysis on these diseases too, in that case, and checking for overlapping pathways (and noting the pathways that differ).

[Md. Al - Mustanjid]

Any source or something I could read and understand? Sorry to say as I am just a beginner.

[Kevin Blighe]

For this exact process that you are doing? - I am not aware of any. Which do you normally use for pathway analysis?