This is a test version of Biostars. For the public version, visit https://www.biostars.org.
Different results between papers and GEO2R

Hi,

I performed a DE miRNAs analyses via GEO2R for https://www.ncbi.nlm.nih.gov/geo/geo2r/?acc=GSE78775. I obtained an upregulation for miR-148a in tumors compared with normal. When I search miR-148a in Pubmed, unfortunately a lot of paper say it is downregulated (in exp., https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4982598/). Could you help me to interpret the GEO2R results or re-perform the analyses according to your codes / softwares, please?

Thanks

microarray geo2r
• 2,310 views • link
updated by ATpoint ATpoint • written by cagdas • cagdas

May sound trivial but did you confirm the comparison order of the classes/groups in your analysis?

• 0 views • link
written by GenoMax GenoMax

Geo2R has an option to print the R code. Please paste that here, probably switched reference levels.

• 0 views • link
written by ATpoint ATpoint

I did it, and obtained same results.

• 0 views • link
written by cagdas • cagdas

Yes, I confirmed it, no problem...

• 0 views • link
written by cagdas • cagdas

There is something strange happening. You may have to contact the authors:

library(Biobase)
library(GEOquery)

# load series and platform data from GEO
gset <- getGEO("GSE78775", GSEMatrix =TRUE, getGPL=FALSE)
if (length(gset) > 1) idx <- grep("GPL10850", attr(gset, "names")) else idx <- 1
gset <- gset[[idx]]

rownames(gset)[grep('miR-148a', rownames(gset))]
[1] "hsa-miR-148a"  "hsa-miR-148a*"

data <- data.frame(
  Tissue = pData(gset)[,'tissue:ch1'],
  miR148a = exprs(gset)[grep('miR-148a', rownames(gset)),][1,])
head(data)
           Tissue  miR148a
GSM2076383  tumor 7.218001
GSM2076384 normal 5.356292
GSM2076385  tumor 9.058600
GSM2076386 normal 6.050635
GSM2076387  tumor 7.393683
GSM2076388 normal 5.517655


par(mfrow = c(1,2))
boxplot(miR148a ~ Tissue, data, main = 'miR-148')
data <- data.frame(
  Tissue = pData(gset)[,'tissue:ch1'],
  miR148a = exprs(gset)[grep('miR-148a', rownames(gset)),][2,])
head(data)
           Tissue     miR148a
GSM2076383  tumor 0.053578910
GSM2076384 normal 0.026429820
GSM2076385  tumor 0.032966353
GSM2076386 normal 0.002947842
GSM2076387  tumor 0.014932436
GSM2076388 normal 0.025117474

boxplot(miR148a ~ Tissue, data, main = 'miR-148*')

kkk

• 0 views • link
written by Kevin Blighe Kevin Blighe

Kevin thank you for your detailed reply.

• 0 views • link
written by cagdas • cagdas

Not for me to follow-up - I would clarify with the authors. This miR is clearly up-regulated in tumor, judging by the box-and-whiskers plot (above). Perhaps the authors are segregating by sex|gender, though (?)

• 0 views • link
written by Kevin Blighe Kevin Blighe

They may have made a mistake with their metadata - not sure.

• 0 views • link
written by Kevin Blighe Kevin Blighe

Even in their published work (Paper), they stated that this miRNA is down-regulated in GC patients. I believe they mixed the samples between tumor and normal when uploaded to GEO . As the others mentioned, best to check with author.

• 0 views • link
written by fernardo fernardo

0 answers

No answers yet.

Log in to answer this question.

More posts like this