Hi. I searched some paper and saw they studied lncRNAs by polyA selected RNASeq library. Is that a reasonable way to do that?
As I saw on wikipedia, lncRNAs are not always with polyA. So if we use polyA selected RNASeq library to analyze, we will miss many lncRNAs, right?
Another question, I found in UCSC table browser, there is a repository for human lncRNAs (lincRNA transcripts, about 30,000, in hg19 version), how I can know are the lncRNAs with polyA or not?
Further, the reason I asked this question is because I have some non-polyA RNASeq in several samples, and have polyA RNASeq in other samples, I doubt if I can compare the lncRNA expression between the different libraries. I know the design is not good, but they are only available resources.
Thanks.
Hi. Comparison would be very tricky between both library types. Especially because you have the chance that one of the two libaries is "more complex", meaning has more different transcripts expressed so all your RPKM values will be different. One thing you could check is just do a basic differential expression on the non-polyA and the polyA+ libraries and check what the top differentially expressed transcripts are. But in the end if you do differential expression between the libraries you will be always biased. You could make some kind of complex design with covariates but you potentially miss a lot of transcripts then...