DIABLO (N-integration) for different omic data and same set of genes

Analysis
1

Dear all

I have two cell lines (control and knockout) for which I have RNA-seq, ChIP-seq, TT-seq and other types of data (in total 8 variables) that I would like to integrate. All 8 sequencing experiments provide information for the same set of genes, although the type of information provided is different (levels of expression, levels of methylation, rate of transcription, etc).

  • Can I use mixOmics to integrate my data?
  • Should I apply a N-integrated approach? I am not sure about this since all my omic experiments provide info about the same set of genes (i.e. all the p or features are the same across all my matrices).

Thank you very much in advance.

2

hi @agallego,
Yes you can use the mixOmics N-integration tools for that (PLS and DIABLO). It will probably either (hopefully) select the same genes across the data sets, or complementary information.

It might be good first to filter your data first as the number of features in each data set must be very large. I.e we recommend the top 5,000 features with the most variance (for example from the RNA-seq, and match them across all the other -seq; or just the most variable features independently across each data set if you are looking for complementary information).

Kim-Anh