How to analyse GSE Series with Multiple Platforms using R?

Question

[Prateek]

Hi, I want to perform differential expression analysis of a GSE Series GSE138214 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE138214) using limma in R. I am unable to extract data from this series as this series has around 14 different platforms.

Using GEOquery and limma package, I am able to differentially analyse genes in Series GSE138260 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE138260), since it has only one platform.

But the GSE138214 series has 14 platforms and I am not sure how to work around it.

Thanks

Answer 1

[Sean Davis]

It looks like you are interested in differential gene expression. Note that there are only two samples, so statistics won't work for you here. Also, these are tiling arrays, not traditional gene expression arrays. Finally, the GEO submission did not include values, so getGEO will have to be augmented with custom code and accessing supplemental files. In short, it will be a lot of work for potentially little reward.

[Prateek]

There are although in total 28 samples (14 AD and 14 Control). But I get the reason why it will be a lot of work with a lot of codes and accessing supplemental files.

Thanks for your reply!

[Sean Davis]

Close, but there are only 2 biological samples, each run on 14 platforms. In essence, each biological sample was prepared once and then run on 14 different arrays, each of which measure independent expression across the genome (not measuring gene expression directly). So, there is no replication in this dataset.

[Prateek]

Thanks for the explanation!