How To Design A Long Noncoding Rna Detection And Quantification Experiment

Question

[kanwarjag]

I am planning on an experiment for detection of novel linnoncoding RNA as well as expression of known lincRNA in patient samples. I was wondering what will be the best approach to extract non coding RNA. I hav eseen Ribo depleted, Ribo zero Hexamer. I sthere a study comparing existing methods of studying such non coding RNA or suggest any best method. I would appreciate if some one in my situation can share their vies, what they have tried and their working experience on which approach one may be best. Thanks

[Zev.Kronenberg]

Is your question bioinformatic related?

[Michael]

I think seqanswers would be a better place to ask this. In my opinion we can leave it open because we are nice (and some bioinformaticians might know more about sequencing than a lot of biologists), but otherwise the question is obviously not sufficiently related to bioinformatics, other than that the analysis implies that bioinformatics tools will be need.

[kanwarjag]

I believe researchers may have dealt with it before analysis and can provide better perspective than just wet lab people. However I do agree that it is not a pure programming question Thanks

Answer 1

[biorepine]

Based on recent studies, it is good to use polyA+, long reads (>70 bp), strand specific RNA-Seq to identify lncRNAs. Briefly, you can follow these steps.

1. De novo assembly of your RNA-Seq with out using known genes annotation after mapping.
2. Remove assembled de novo transcripts that overlap any known transcripts.
3. Calculate coding potential by scanning conserved ORFs (open reading frames) and remove the ones which look like coding.
4. To further strengthen your lncRNA, overlap H3K4me3 peaks with promoters of lncRNA and H3K36me3 peaks with the gene body.
5. Remove transcripts with less than 2 exons
6. Remove transcripts that are smaller then 200 bp