Hello,
I would like to utilize a published dataset of RNA-Seq for extracting information, such as upregulation/downregulation of transcripts of relevance to my project. The published dataset is from wildtype mice, untreated and uninfected. I would like to use this to look for transcripts of my interest just for preliminary data acquisition. What kind of tools/analysis/workflow would be suitable for this?
Thanks for your help!!
The easiest way to extract the data is generally the SRA-toolkit. Sometimes the authors also upload some expression data along with their project files, so you might be able to use that directly.
Most likely, you will have to repeat their analysis. This involves alignments (HISAT2 is ideal), along with some sort of differential expression analysis tool (cuffdiff is a bit outdated, DESeq2, BallGown on bioconductor). Most of these require a basic to intermediate familiarity with linux or R; so as others have suggested you may consider using Galaxy. Although I have found it to be slower and less flexible than the command line, it is very user friendly.
If you want to skip the alignements / counting steps, I would recommend to look for a study of interest on GEO or SRA (NCBI websites) and then look in the EMBL RNASeq-er database (https://www.ebi.ac.uk/fg/rnaseq/api/) to see if they have the aligned data (more than 350,000 RNA-seq datasets). You can search by GEO or SRA study number. Then you'll probably have to use some differential analysis tool (I would recommend DESeq2), if you want to discriminate some genes which are up- or down-regulated in one situation compared to another, if not you can just have a look at a basal expression level in untreated wild-type mice to get an idea. I agree with Michael above on the use of Galaxy (https://usegalaxy.org), especially if you're not at ease with R / bioinformatics. Good luck!
Well, you will be able to utilize from the too much raw data sets whose been deposited in ArrayExpress and GEO. By contrast, for the obtaining up- and down regulated data in thus you are obliged to looking for at additional files in published articles.