Integrating Mutation, RNAseq and RPPA data toolset

Question

[pb11]

Hello All

I have dataset of 60 samples with its corresponding RNAseq and RPPA. I want to integrate this data with some mutation data that I already have for these 60 samples. I am looking for some suggestion on which toolset should I use to integrate these datas. I want to generate a plot where I can put all these dataset together. Any suggestion??

Thanks in advance

Prit

[ivivek_ngs]

Depends on what you want to achieve. Do you want to use integration via 1. correlation or probabilistic modeling? 2. Use the unsupervised method (MOFA) or supervised method (mixOmics-DIABLO)? 3. Simple canonical correlational analysis (CCA) using mixOmics and then look up the gene sets between RNASeq and RPPA to understand if there is an upstream mutational evidence available for those genes?

Is your aim only to integrate the datasets and plot the variance? What are the phenotypic categories that you want to study? Do you have longitudinal data or cross-sectional data? Generating a plot can be either done in a lot of ways and does not even need any tool but if you want to used established tools then definitely the below links can provide the start.

For MOFA

For mixOmics

For mixOmics-DIABLO

Check out this review Figure 1 to find a list of available tools for multi-omics integration as I understand you have mutation data, RNASeq, and RPPA from samples. Good luck.

[pb11]

Thank you for the help. Let me give you 1 examples: I want to make a representative graph where I show p53 mutations (whole genome), its downstream pathway (RNA seq and RPPA) in 1 figure. How should I represent that?

[ivivek_ngs]

Well, this is a very different question than what you asked for in this thread. This is a hypothesis that you want to present in a graphical abstract.

1. Have you already used any of the multi-omics integration tools at your end to see if your data already shows a p53 mutation that regulates a set of genes from your RNASeq and RPPA data?
2. If so, do these gene sets enrich for p53 mediated signaling pathways or networks?

If you find the above you can definitely graphically represent them using a network map (e.g. Cytoscape or any available open-source tools). But such network creation is beyond integration of the omics data and query of your thread. It will be more after you integrate data and perform a downstream functional analysis of your integrated results. Hope this gives you some lead.