I am completely new to meta-analysis and am a bit lost on how to proceed.

I have two biomarker datasets; each one is using a different microarray platform. I have the protein names, the concentration in each subject (one dataset has 15 subjects while another has 22) as well as fold change, p-value, and AUC values.

Is there a simple way to do a meta-analysis with this information? For example, is there any way to join the datasets and calculate a combined fold change and AUC? I was thinking of using sumlog() from the metap package in R to do a combined p-value, but I am lost on how to combine the other two values, or if this would even be appropriate. I've attached a screenshot of one of the datasets

dataset preview

Any help or guidance would be greatly appreciated!

  • 1
    $\begingroup$ You seem new to analyze microarrays. Look at the wonderful limma vignette about how to analyze one dataset. Then I suggest you to analyze both datasets and see how much do they agree. However what you need to do depends on what you want to do (besides a meta-analysis). What is your biological question (or reason of this analysis)? $\endgroup$
    – llrs
    Commented Nov 3, 2020 at 12:19
  • $\begingroup$ Thats a lot of questions rolled into one. First consider data mining to get an overview $\endgroup$
    – M__
    Commented Nov 5, 2020 at 14:46


Browse other questions tagged or ask your own question.