hello, allen institute.
I am fairly new to working with transcriptomics data. I download human brain data from Microarray Data Download :: Allen Brain Atlas: Human Brain. I want to do some differential gene expression analysis between diffrent brain regions using ready-made R package such as “limma”. But i did not find any apparent description of the data format. In MicroarrayExpression.csv, data is count (after log2)? or TPM,FPKM (after normalization)?
Thanks!
Hi @jiwieli. We have a white paper describing how the data was normalized (Microarray Data Normalization), but essentially each number is a value representing the amount of light emitted from the indicated probe on a specific microarray chip. These values are quantitative (e.g., you can believe that a two-fold change in value really means a two-fold change in gene expression), but don’t represent absolute expression levels in the way that you get with RNA-seq data. The values in the downloaded table are log normalized. There are a few other Community Forum posts on this topic (type “Microarray” in the search box) if you are interested in more information.