NuGEN or Truseq RNA-seq system?

[MouseCrusader]

I am mapping out my journey into the RNA-seq world and would appreciate additional perspectives on the decision to use NuGEN or Truseq technologies for multiplexed RNA-seq on the Illumina HiSeq platform. I will be characterizing the transcriptome of a small subpopulation of cells that will be a challenge to acquire the 100ng of total RNA material that is suggested as a minimum for Truseq sample preparation.

Does Truseq prepared 100ng total RNA provide a much richer representation of the transcriptome than 10-25ng total RNA, that is more within reach, prepared with NuGEN Ovation system?

[elaney_k]

Hi,

I've used NuGEN RNA-seq sample prep reagents and would never recommend them to anyone. The resulting data looked awful with far more reads mapping to intronic regions over exonic regions. The reads did not correspond with any non-coding RNA species at all and no satisfactory explanation from Nugen was forthcoming, other than to say that the data looked very similar to the data they get in-house.
The same RNA samples were then prepped with the Illumina DSN protocol and yielded far superior and logical results.
Just my 2c but never again will I try a NuGEN product!
Elaine

[MouseCrusader]

Hi Elaine,
Thank you for sharing your experience. What was your input total RNA amount you used for NuGEN and Illumina DSN systems?
Congratulations on finding success!
Dan

[elaney_k]

Hi Dan,

Sincere apologies for such a late response! We used 50ng of globin reduced total RNA as input for the Ovation RNA-seq protocol and 2ug of total RNA as input for the Illumina protocol (which underwent the standard polyA purification protocol before the DSN treatment).
The RNA samples were from human whole blood.

Elaine

[Simon Anders]

Are you aware of this paper here that appeared last week in NAR? They compare NuGEN and TruSeq.

Muhammad A. Tariq, Hyunsung J. Kim, Olufisayo Jejelowo and Nader Pourmand:
Whole-transcriptome RNAseq analysis from minute amount of total RNA
Nucl. Acids Res. (2011), doi:10.1093/nar/gkr547

[NextGenSeq]

Yes, similar to the paper referenced we have obtained excellent results with the NuGen Ovation system.
Note make sure you DNase treat your RNA or else your library will have a lot of mitochondrial DNA in it.

[pbluescript]

I have also used the Nugen kit, and I am pleased with the results. The inputs I had were less than 1ng and degraded, leaving me with little choice in how to prepare a library for NGS.
More than half of the reads mapped to intergenic regions, but on a single lane of the HiSeq, it's still plenty for much of what I want to do. It's essentially a problem you can buy your way out of if you have difficult to obtain or unique samples. My follow up has gone well with qPCRs matching the predicted changes seen in FPKM values and novel isoforms and RNA edits confirmed with other methods.
If you can't get the RNA amount up high enough, I'd definitely recommend the Nugen Ovation kit.

[MouseCrusader]

Thank you all for sharing your experiences and information. A colleague's laboratory is currently test-driving the NuGen kit with low sample input and will know soon if they are able to have success with the system, which will in turn be a good resource for my own library preps.

[PigletIA]

Hi Dan,

Sincere apologies for such a late response! We used 50ng of globin reduced total RNA as input for the Ovation RNA-seq protocol and 2ug of total RNA as input for the Illumina protocol (which underwent the standard polyA purification protocol before the DSN treatment).
The RNA samples were from human whole blood.

Elaine

Hi, Elaine,
I’m also thinking about using DSN to treat polyA selected RNA-seq library (using Truseq kit). Do you think you could answer/comment on some of my questions about the detail of this approach?
1. Do you introduce the DNS step after the 2nd strand synthesis (before add the adaptor) or after the final PCR amplification?
2. Did you use the Illumia recommended DSN condition?
i. Final sample library: 80-100ng
ii. Buffer (final 1x): 50mM HEPEs+500mM NaCl
iii. Melt library at 98Cx2min
iv. Hybridize library at 68Cx5h
v. DSN cleavage: 68Cx25min
vi. PCR to enrich library
Thanks a lot!
Lan

[CC_seqanswers]

Hi, Bluescript,

Thanks for your post. We have had good experience with Ovation kit too.

One thing I don't quite understand is why more than half of reads mapped to intergenic regions with ovation kit. I would think the intergenic reads came from DNA contamination and it should not have much to do with whichever protocol used for library prep. Or I might be wrong?

CC

I have also used the Nugen kit, and I am pleased with the results. The inputs I had were less than 1ng and degraded, leaving me with little choice in how to prepare a library for NGS.
More than half of the reads mapped to intergenic regions, but on a single lane of the HiSeq, it's still plenty for much of what I want to do. It's essentially a problem you can buy your way out of if you have difficult to obtain or unique samples. My follow up has gone well with qPCRs matching the predicted changes seen in FPKM values and novel isoforms and RNA edits confirmed with other methods.
If you can't get the RNA amount up high enough, I'd definitely recommend the Nugen Ovation kit.

[pbluescript]

Hi, Bluescript,

Thanks for your post. We have had good experience with Ovation kit too.

One thing I don't quite understand is why more than half of reads mapped to intergenic regions with ovation kit. I would think the intergenic reads came from DNA contamination and it should not have much to do with whichever protocol used for library prep. Or I might be wrong?

CC

I don't understand the intergenic reads and Nugen doesn't seem to have a clue either. I've only seen this effect with cDNA made from the Ovation kit, and not from other methods. I do notice that the reads seem to be clustered rather than randomly distributed. I will see hundreds of reads mapping to a small location and then nothing for long stretches. I might be wrong, but I would assume if it was gDNA contamination, the pattern would be a bit more random than that. I have been focusing on my reads that map to genes, but it might be a good idea to see if the intergenic reads coincide with any sort of repeat elements.

[pmiguel]

I don't understand the intergenic reads and Nugen doesn't seem to have a clue either. I've only seen this effect with cDNA made from the Ovation kit, and not from other methods. I do notice that the reads seem to be clustered rather than randomly distributed. I will see hundreds of reads mapping to a small location and then nothing for long stretches. I might be wrong, but I would assume if it was gDNA contamination, the pattern would be a bit more random than that. I have been focusing on my reads that map to genes, but it might be a good idea to see if the intergenic reads coincide with any sort of repeat elements.

The answer is in the manuscript posted up-thread:

Muhammad A. Tariq, Hyunsung J. Kim, Olufisayo Jejelowo and Nader Pourmand:
Whole-transcriptome RNAseq analysis from minute amount of total RNA
Nucl. Acids Res. (2011), doi:10.1093/nar/gkr547

See figure 1a. NuGen Ovation performs essentially the same as RiboMinus. That is, it depletes rRNA (somehow...) but does not otherwise select for polyadenylated messages.

So it all comes down to the the composition of your total RNA. Even after you deplete the >80% rRNA from the sample, there are still a wide variety of RNAs remaining -- only a fraction of which are polyadenylated.

If you only care about polyA messages you will be sorely disappointed with all the other "junk" you will see. This is not the method for you. You will need to isolate intact total RNA and use an oligo dT based capture method to focus your results.

--
Phillip

[CC_seqanswers]

It's true that Ovation method is like whole transcriptome sequencing protocol, however， I still don't quite get that >50% reads mapped to intergenic area.

Would it be possible a slight amount of DNA carry-over got amplifed crazy? polyA method works better for DNA contaminted RNA sample as it's a "fishing out" procedure other than"leaving behind" like Ribomius

The answer is in the manuscript posted up-thread:



See figure 1a. NuGen Ovation performs essentially the same as RiboMinus. That is, it depletes rRNA (somehow...) but does not otherwise select for polyadenylated messages.

So it all comes down to the the composition of your total RNA. Even after you deplete the >80% rRNA from the sample, there are still a wide variety of RNAs remaining -- only a fraction of which are polyadenylated.

If you only care about polyA messages you will be sorely disappointed with all the other "junk" you will see. This is not the method for you. You will need to isolate intact total RNA and use an oligo dT based capture method to focus your results.

--
Phillip

[pmiguel]

It's true that Ovation method is like whole transcriptome sequencing protocol, however， I still don't quite get that >50% reads mapped to intergenic area.

Would it be possible a slight amount of DNA carry-over got amplifed crazy? polyA method works better for DNA contaminted RNA sample as it's a "fishing out" procedure other than"leaving behind" like Ribomius

Sure, it is possible.

But perhaps you are laboring under the illusion that the majority of transcription contributes to the pool of poly-A RNA that is ultimately translated? This is not the case. Even after removing rRNA, the majority of RNA is not translated -- it functions in various complex regulatory pathways, many of which are still being elucidated.

Also there are the transposable elements that comprise the majority of many genomes. You might be seeing a lot of RNA from one or many of them.

What organism and tissue are you sequencing?

--
Phillip

[Veleno]

There was years ago, still in the chip chip era, a paper on Science (from Affy perhaps) describing transcription from the majority of the genome; so no surprises there