Biostar Beta. Not for public use.
Analysis past the differentially expressed genes: RNAseq
Entering edit mode
23 months ago
biogirl • 170
European Union

Hi all,

I have a very open-ended question, as I just wanted to gauge what people's thoughts were rather than get a concrete answer. There are a lot of papers coming through now where RNAseq has been used and differentially expressed (DE) genes have been identified, and perhaps some gene ontology analysis has been done. But is there anything else that can be done, anything a bit different, or dare I saw ' _cooler '_?

I realise that one can infer gene regulatory networks from time series data, but is there anything else that can be done for single time point (which is quite common) data?


Entering edit mode
15 months ago
  1. GO analysis (mentioned above) (For all the 3 components)
  2. Pathway analysis \
  3. PPI /Signaling network analysis (IPA is best for this, in my opinion). Sting is also good
  4. Cluster analysis
  5. Immediate promoter analysis collect immediate promoter sequences (1000 bp upstream of TSS) for significant genes and see if they share same transfactors.
  6. Chromosomal location analysis Map the DE genes on their chromosomal locations and see if they cluster (For eg. MHC genes cluster in humans, mice)
  7. Protein motif analysis See if they share any common protein motifs and motif binding factors
  8. miRNA analysis See if they share similar miRNA regulation. Confirm it either by public db or in-house experiments 9) Call variants from RNA seq and see if they match with WES else where. If there are any discrepancies (barring experimental and technical errors). Look for RNA editing events. Not sure if this works. Look for NMD transcripts.
  9. Look at the missense/nonsense/truncated transcripts and model them (homology or abinitio). See if they change any of physical, chemical and functional properties of the protein
  10. Look at the somatic and germ line variants from RNAseq and classify them based on disease,drug target databases. Check which is most represented.drug/disease and compare with your experimental observations.
  11. Look for transcript switching

There are plenty of applications for transcriptomics data (RNAseq, miRNAseq, Exon arrays). I am sure scientists here, will come up with 1000 more analyses.

Entering edit mode

Brilliant suggestions, a very thorough list of analyses, thank you

Entering edit mode
13 months ago
Belgium, Brussels

- fusion gene

- alternative splicing

- lncRNA/lincrna/circRNA detection

- virus detection

Entering edit mode

Thanks for these suggestions - IncRNA is something I really hadn't considered, and definitely falls in the 'cooler' category!


Login before adding your answer.

Similar Posts
Loading Similar Posts
Powered by the version 2.3.1