Hi, I am trying to look at differential protein expression between multiple sample types with biological replicates for each type. Data was generated using nanostring DSP geoMX and normalised in-house. I haven't been able to find an R package to do differential expression/volcano plots from already normalised dataframes, as Deseq2 requires raw RNAseq counts. Just wondering if someone could point me to a package that will do this type of analysis?

Thanks James

In order to understand where to go next with your analysis, you will have to understand more about the data that has been produced by geoMX. I have not used this specific program suite but there are likely different options for how to export the data, e.g., as normalised counts, log2 normalised counts, etc. There may even be options within the program for how to conduct differential expression analysis, or to output raw counts such that you could separately use something like DESeq2 for normalisation and differential expression.

If you have run your own NanoString experiment, though, then you should have access to the raw data RCC files. In this case, there can be 2 approaches:

GUI-based: nSolver

If you are more comfortable using point-and-click GUIs, then nSolver may be the best option for you. It normalises the raw RCC data and can conduct differential expression analysis via the Advanced Analysis plugin (installed separately), and can perform PCA and generate volcanos.

Command-line:

If you are okay with the command line, then you can import RCC data into a raw count ExpressionSet object via NanoStringQCPro, and perform QC using this package, too. After QC, you can normalise the data via RUVSeq / DESeq2, and fit the actual GLM for differential expression via the standard DESeq2 functions.

Another option is NanoStringDiff package.

Kevin