Ion Torrent RNA-Seq v2 - Success?

[david135]

fsdfsdfsdfdsfdsfsdfsdfdsf

[jonathanjacobs]

We are using the RNAseq v2 kit routinely for viral RNA sequencing, miRNA sequencing, and polyA+ libraries too. Works quite well. We did have a Ashe of a bad lot / kit once and life tech replaced the whole kit ($4800) no charge - of course the lost time wasted was annoying.

[david135]

sdfsadasdsadsadsadsa

[david135]

sdasdsadsadsadasdsadsad

[thomasblomquist]

Just finished developing our method for targeted quantitative RNA sequencing. Our studies use Ion torrent platform, but this approach works well on Illumina as well. Our unpublished data shows our results are concordant between the two platforms as well. These latter studies will be out in 2014. These studies are part of a larger FDA initiative for concordance between RNA sequencing studies.

http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0079120

Regards, Tom Blomquist

[snetmcom]

Also is there any data published using Ion Torrent with RNA-Seq?

this kit just an update from the previous versions on SOLiD and Illumina. You can find plenty of those publications.

[Gorbenzer]

Hi guys, we have trying to use v2 kits for Bacterial transcriptimics. All values during library preparation seems good but final output is really low, with 50/60% of polyclonals. It seems that the step generating problems is the onetouch.

Anyone have any suggestion? any alternative protocols?

Thanks to anyone that will help us!

[thomasblomquist]

Hi guys, we have trying to use v2 kits for Bacterial transcriptimics. All values during library preparation seems good but final output is really low, with 50/60% of polyclonals. It seems that the step generating problems is the onetouch.

Anyone have any suggestion? any alternative protocols?

Thanks to anyone that will help us!

Heterodimerization is the primary cause of polyclonal bead population. Which is to say denaturation of complementary homodimer DNA strands, and subsequent reannealing with a somewhat complementary, but not perfect match. This then results in a "polyclonal read" because the sequences are out of frame with each other. This occurs during the downstream PCR amplification step if too many PCR cycles are used, too much starting cDNA is used (early plateau of PCR and loss of exponential phase product formation), or during subsequent purification steps if the sample material is denatured because of heat or chaotropic salts, etc. Thus, all goals should be optimized for keeping the "Homodimer" DNA products after second strand cDNA amplification matched up. Else, you get polyclonal reads.

This can be frustrating.

Check out the methods section in our paper, and we describe our approach to avoiding the heterodimerization issue which leads to polyclonal reads.

http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0079120

-Tom Blomquist

[bioits]

Hi guys, we have trying to use v2 kits for Bacterial transcriptimics. All values during library preparation seems good but final output is really low, with 50/60% of polyclonals. It seems that the step generating problems is the onetouch.

Anyone have any suggestion? any alternative protocols?

Thanks to anyone that will help us!

Do you have BioA traces for input RNA and cDNA? If everything seems ok, I would try to dial down the input amount for OneTouch.

[mbzmg1]

We have about 30 successful RNAseq runs on the Ion Torrent Proton using the RNAseq v2 kit with polyA selected RNA. One modification we have made is to size select a narrower range (110-200bp) using the Pippin system. We have found that this results in a higher "effective" library concentration when using the Bioanalyzer quantification range as suggested in the RNAseq kit manual (ie more of our final library is in the optimal size range for sequencing). We had to lower the concentration from the recommended 11pM to ~5pM to keep the polyclonal % under control and have been able to get up to 88M reads (average is more like 72M). The recommended concentration has been changed in the OT2 200 v3 manual but we have lowered our input accordingly and are getting similar results.
I would suggest lowering your concentration by ~0.5-1pM each run and monitoring from there.

[NextGenSeq]

We find miRNA-Seq works very well (actually MUCH better than Illumina).

We find that NuGen works better for standard RNA-Seq. In comparison experiments between Illumina RNA-Seq and Ion Torrent RNA-Seq we got correlation coefficients of 0.998

[RemitoAmigo]

Hi guys,

We have about 10 successful run for RNA seq on the proton. We use the RNA-seq v2 kit from Life tech. Both Ribo- and Ribo gold and RNaseIII fragmentation and (chemical from NEB). We also have the AB lib builder...which was not so successful at first but getting better....but 13 libs in a batch is a little slow.

Our polyclonality is in average about 30%...

MBZMG1 what are you getting for polyclonality?...we are sequencing with the V3 kit.
NEXTGENSEQ can you provide more info about the NuGen library kits (links and what metrics are you using to define the NuGen kit as 'working better')?

Tx.

RemitoAmigo

[mbzmg1]

MBZMG1 what are you getting for polyclonality?...we are sequencing with the V3 kit.

Our polyclonal percentages are in the low to mid 20s with V3 OT2 and SEQ kits.

On another note -what are people using for alignment? We have found that Bowtie2/Tophat2 do not play well with Ion Torrent data and have had much better success with GSNAP.

[RemitoAmigo]

I'm using STAR + bowtie right now and getting 75% alignment on mouse ref. But I'll definitively give a try to GSNAP.

Tx.

RemitoAmigo

[Brian Bushnell]

I think you'll find that BBMap works well for Ion Torrent RNA-seq data; it is more tolerant of indel errors than any other RNA-seq aligner. The post I linked to contains an example command line.