Truseq RNA bioanalyzer profiles

[Swathi]

The libraries were prepared using Truseq RNA sample prep kit V2. I want to know why the library profiles do not look the same for both the samples. Why is one of the profile more skewed at 300bp whereas the other profile is flat.

Thanks

[nucacidhunter]

None of the libraries looks normal for the kit. The left one has slightly more large fragments and the right one a lot of large inserts. A possible explanation is under-fragmentation.

[Swathi]

As per the kit specification, for insert length range of 130-290bp 2 mins of fragmentation at 94 degree Celsius is recommended. The same was carried out for these libraries too.

[nucacidhunter]

If fragmentation step was incubated for 2 min then the left one is ok. The other one would have had some issues in reaction set up or input material. Perhaps it had some DNA contamination or RNA was eluted in buffer which was not optimum that has affected fragmentation.

[mayankkaashyap]

Hi,
I have fragmented ribo-depleted RNA for 2 mins (pic attached). Is it fine if I want to have average 400 bp size library. I am using TruSeq Total RNA stranded kit and purposely want to sequence larger inserts lesser than 500 bp.

[adam.geber]

mayankkaashyap, if the trace you attached represents your fragmented RNA then it definitely won't be possible to make a ~400bp library from it without further fragmentation. What conditions were you using after the ribo-depletion?

Edit: I was incorrect with my interpretation -- cDNA synthesis should help reduce the average fragment size, as nucacidhunter says below!

[nucacidhunter]

Hi,
I have fragmented ribo-depleted RNA for 2 mins (pic attached). Is it fine if I want to have average 400 bp size library. I am using TruSeq Total RNA stranded kit and purposely want to sequence larger inserts lesser than 500 bp.

According to kit user guide, 2 min fragmentation would result in a library with insert size in 130-290 bp. The reason is because during 1st strand synthesis RNA will be primed randomly and large fragments are more likely to be primed from multiple locations which will reduces the size of ds cDNA and consequently insert size of final library. Even if you skip the fragmentation the largest insert would be 350 bp. If you want larger insert you may need to use other kits where the full length ds cDNA is synthesised and then it follows a standard DNA library prep workflow.

[mayankkaashyap]

Hi nucacidhunter,
I used no fragmentation step (attached) and Illumina analyst said it is not advisable to proceed with sequencing coz the inserts were large and cause problem on sequencer. I plan to do 1- 2 min fragmentation so as to get most of the inserts in range of 200 bp - 500 bp, even if there is random priming. Even there will be coverage risks more than 500 bp inserts are there

[nucacidhunter]

Hi mayankkaashyap

Your sample 9 and 11 main peak is around 400 bp and that translates to 280 bp peak insert size. This is according to manual with no fragmentation. I mentioned priming because your fragmented RNA length attached to post #5 was up to 6 kb to emphasize that RNA fragment length does not translate to library insert size. If you want to do expression analysis the recommendation is for 9-8 min fragmentation which gives an average insert size of 150 bp which gives more even coverage. If you are looking for other features such as splice variants, then you probably have to focus on that rather than simultaneous expression and splicing event identification. I hope someone in the forum with more experience in bioinformatics can comment on this if that is what you are looking for.

Large fragments does not cause sequencing problem, but they are less likely to form clusters and hence do not get sequenced. By large I mean fragments over 1 kb. Sequencing a library with wide size range results in reads with smaller insert peak than library peak because shorter fragments are more likely to cluster than larger ones.

[mayankkaashyap]

Please advice me. I want to do long non-coding gene expression analysis. In plants it is reported between 200 bp to 500 bp (Bioanalyser bp). What will be best for me to do.

[nucacidhunter]

I would suggest you start a new thread in Bioinformatics Forum with a title such as: How to analyse expression of long non-coding genes using RNA-seq

I do not think you need a large insert RNA-seq library for your study aim. It will be also useful if in the new thread you can provide more information about species genome size, availability of reference genome or transcriptome.

[mayankkaashyap]

Hi nucacidhunter,
So I decided to do 8 min fragmentation. Here is final library BA trace (attached). My worry is the second peak at about 700 bp. What should I do??

[nucacidhunter]

It looks like overcycling PCR. I wonder what was input and PCR cycle numbers. We normally get the peak at 300 bp without the larger one.

If it is due to overcycling it would not cause any trouble for sequencing but following should be considered:

1- Quantify library only with qPCR (Qubit or PicoGreen results will not be reliable)
2- Only consider small peak average size for calculating molar concentration ignoring the large peak

[mayankkaashyap]

Thanks for your reply, nucacidhunter.
Input was 1 ug Total RNA and I used 15 PCR cycles. I will use qPCR to dilute molarity.

[nucacidhunter]

I have not seen a profile like this by 1ug input and 15 PCR cycles. I suspect that you are using half reaction reagents. Anyway sequencing would be fine. But keep in mind that it might affect expression results.