Questions about RNA fragmentation?

[zhzliang]

I'm almost crazy. Recently, I've prepared two batches of libraries for mRNA-seq. Bioanalyzer analysis all show the two batches of libraries have the peak sizes between 180~190bp (means the inserts shall be 60~70bp, Pair-End). Although the first batch looks good after analysis. But my boss severely concerned about the short insert length. I just do as instructed and got the similar peak size (180-190bp) as the Illumina mRNA-seq protocol.
What I don't understand why the insert size is so short? And many people in this forum have mentioned the optimal size of insert shall be more than 120bp, even up to 500bp. Adding the adapter/primer length, the overall library size shall be larger than 240bp. But why I only see a 190-200 peak size in the Illumina mRNA-seq protocol? I believe the 200-peak-sized library can be sequenced. But I can't explain where the 40bp difference has gone. Could anyone give some suggestions about this? Can these 180-190bp-peak-sized libraries be sequenced?
Another question. After fragmentation, the RNA looks smear in RNA denaturing gel with wide size distribution. How to get uniform size of RNA fragmentation without gel purification?

[shurjo]

I'm almost crazy. Recently, I've prepared two batches of libraries for mRNA-seq. Bioanalyzer analysis all show the two batches of libraries have the peak sizes between 180~190bp (means the inserts shall be 60~70bp, Pair-End). Although the first batch looks good after analysis. But my boss severely concerned about the short insert length. I just do as instructed and got the similar peak size (180-190bp) as the Illumina mRNA-seq protocol.
What I don't understand why the insert size is so short? And many people in this forum have mentioned the optimal size of insert shall be more than 120bp, even up to 500bp. Adding the adapter/primer length, the overall library size shall be larger than 240bp. But why I only see a 190-200 peak size in the Illumina mRNA-seq protocol? I believe the 200-peak-sized library can be sequenced. But I can't explain where the 40bp difference has gone. Could anyone give some suggestions about this? Can these 180-190bp-peak-sized libraries be sequenced?
Another question. After fragmentation, the RNA looks smear in RNA denaturing gel with wide size distribution. How to get uniform size of RNA fragmentation without gel purification?

There are a couple of levels of complexity in your question, so let me take a top-down approach and see if I can help you here.

1. You can certainly sequence a library where the insert size (inner distance between the PE pair) is 40 bp. However, the longer your insert size is (upto a certain point), you are going to have more power to do certain types of analysis in the downstream parts of your project (splice junction discovery/transcript assembly/isoform abundance estimation). As such, it would NOT be preferable to repeatedly implement a protocol with an insert size that small.
2. As regards getting a uniform size distribution for RNA fragmentation, I have had great success using the Covaris platform. I previously posted on this forum with my machine settings about a year ago when I was standardizing the protocol. If you cannot find those posts let me know and I will dig them up. Other than that, the Illumina kit probably still uses the Ambion buffered zinc acetate reagents which I had lots of trouble with because the fragmentation peak was never reproducible.

HTH,

Shurjo

[seqseq123]

mRNA fragmentation

Before you deviate from using the Illumina kit reagents, do this:

1. Please ask Illumina to replace the reagents that did not work for you. With right kind of evidence, they will gladly do it for you.

2. Are you sure you had mRNA confirmed by the Bioanalyzer?

2. Take some extra mRNA you have (or prepare some more, or buy some HeLa mRNA) and fragment it with variable temperature. Say, 95C for 3 minutes, instead of 5 minutes.

3. If possible, use 5ug of total RNA for Isolating your mRNA. That will yield sufficient quantity of mRNA that would nicely withstand 3 minute fragmentation time. Add the Illumina Frag buffer (not anything else!!!!) and place the sample in the thermocycler as soon as possible. Remove after three exact minutes on to very cold Metal block- then add the Stop solution as soon as possible, and vortex and cool it on the ice. Now you can go for overnight precipitation.

After you make the double stranded cDNA, you can do the bioanalyzer again and confirm that you have cDNA with reasonable size distribution, and now you can proceed to ligating the adapters.

Hope this helps.

[GKM]

I wouldn't be concerned at all by the fragment lengths you are getting, unless you specifically want to have very long inserts for some reason. However, for your average RNA-Seq experiment, insert sizes of 500bp are very bad idea as you start getting very uneven coverage and you begin to underrepresent shorter transcripts. 200bp fragment size (50bp inner mate distance for 2x75 reads) is about optimal in my experience

[zhzliang]

Thanks everyone for your valuable experience. I need try more to get deep understanding of RNA-seq.

To GKM, to get 200bp inserts, do you mean I need to cut a 320bp gel band after adapter ligation?

[GKM]

When I refer to insert and fragment lengths, I mean the original piece of RNA, i.e. without the adapters.

[zhzliang]

When I refer to insert and fragment lengths, I mean the original piece of RNA, i.e. without the adapters.

Thanks!

I'll try to do like your said.

[EADIE]

Rfam

I have some data of Arabidopsis for small RNA analysis. Now i want to do a locall alignment with Rfam then exclude the sequences which match to Rfam from futher analysis. But the question is:
Do i need to extract the relevant Arabidopsis members of the Rfam before alignning? If it's best to do that, how should i deal with members with no species name? Do they also need to extract for matching? These members are as follows:

>RF00005;tRNA;AAVX01183092.1/735-806 :
GTCCCATAGCATTGTGGTTTGCGTATTCGCCTTGCAAGCGAGAGGATCTG
GGTTCGATTCCCAGGTGGGGCA
>RF00005;tRNA;AAVX01300616.1/338-420 :
GTAGTTGTGGCTGAGTGGTTAAGACAATGGACTAGAAATCCACTAGAAAA
CTACCCGCGCAGGTTCGAATCCTGCAGACTACG
>RF00005;tRNA;AAVX01004056.1/1665-1747 :
GAAGTATTGGCCGAGTGGCGAAGGCGATGGACTACAAATCCATTGAGAAA
CTTCCCGCGCTGTTTCGAATCCTGCCGACTACG

Looking forward to your reply~~
Thank you in advance!