Dear Vahid,

Take a look at the article:

https://www.biorxiv.org/content/10.1101/675777v1.full

Thank you!

Hi Vahid,

How were the sequencing libraries prepared for your samples? If you have Total RNA sequencing data (RNAseq), to capture the trascriptome, a Poly-A selecting library (like the "classic" TruSeq RNA by Illumina) could have been used. In this case you'll only have mRNA and polyadenilated lncRNA.

There are less stringent library (e.g. TruSeq Stranded Total RNA library), with no selection or size exclusion. This case would be better for you as in addition to mRNA and poly-A lncRNA you would also have non-polyadenilated RNA and few other non coding RNAs, probably snRNA and snoRNA.

From these types of libraries you won't be able to effectively identify miRNA or other small RNAs (piRNAs, siRNAs...) because of their small size (most of them should be lost during purification and library preparation) and because the sequencing is heavily biased towards long, abudant sequences. You may find some "residues" miRNA, but not enough to have solid data.

To analyze these type of data i suggest you to use HISAT2-Stringtie-Ballgown instead of TopHat2 (which is sort of deprecated). http://daehwankimlab.github.io/hisat2

HISAT2 is easy to use and pretty similar to tophat and there are a lot of pipelines for lncRNA profiling in the literature.

As for identifying miRNAs, you would need small RNA sequencing, prepared with specific libraries that select molecules by size (excluding mRNA and various lncRNAs). After you clean your smallRNA sequencing data you can directly compare the sequences with miRNA databases or align them with Bowtie, sRNAmapper (for example) or use tools to find the precursor.

I may be wrong, but the best solution to check miRNA and lncRNA in the same sample would probably be to extract total RNA (e.g. with trizol) and then prepare two libraries, a smallRNA one and a total (long) RNA one and sequence them. This way you have your small and long RNA data from the same sample. If you do not have enough RNA to prepare two indipendent libraries you can also use a fractional protocol to divide molecules by size and prepare the library only from small and large RNAs (doable, but not very efficient anyway).

If you already have only smallRNA or RNAseq data I don't think you can retrieve lncRNA and miRNA (respectively) from them to have reliable data.