Differences in coding:UTR read distribution between library preps

Question

[dzb]

Does anyone have an idea about why I would see differences in the proportion of reads mapping to coding regions vs. the 3' UTR in one library prep method vs. another? They are both poly-A purification based but there's a significantly greater proportion (~15%) of reads aligning to the 3' UTR in one library prep method vs. the other. Both methods attribute roughly 80% of all reads to transcripts (non-intronic or intergenic), just one has more in the UTR than the other. Not sure why this would be. Ideas are appreciated!

[WouterDeCoster]

Was the RNA of the same quality? You could have a more pronounced 3' mRNA bias profile due to degraded RNA.

[dzb]

That's a good question and idea. I don't have the information regarding the RNA quality...I'll have to track down this info from the people that QC'ed the RNA extracts.

[Carlo Yague]

That kind of difference could also come from the library prep methods you used. For instance if the cDNA synthesis is based on the hybridation of random primers, you could get less reads in the 3'UTRs compared to method that involve direct adapter ligation on the RNA.