Metatranscriptome analysis - MiSeq or Ion Torrent?

[ramsemm]

Hi Everyone,

I am interested in sequencing the metatranscriptome of ~8-16 soil samples, and I have the option of either having the libraries prepped for an in-house MiSeq run or prepped and run on an Ion Torrent.

Having no previous experience with cDNA libraries or Ion Torrent runs, I was hoping to get feedback from the more experienced as to which platform may be better for my application with regards to read quality, amount, and overall cost.

Please let me know if there is any additional information I can provide. Any feedback is appreciated.

Thanks in advance!

[Jean]

Hi,
We have an Ion Torrent and the output is not really high enough for a metatranscriptome. Especially since you won't be able to multiplex the cost will skyrocket quickly. The data has a lot of the same issues as 454 (homopolymer "stuttering", and poor quality ends) so alignment will be more of a challenge.

I'd recommend the Miseq for sequence quality, sequence output, and lower cost.

[epistatic]

I'd run these type of samples on a MiSeq. Sequence the V4 region of 16S by amplifying using primers 515F and 806R, run libraries PE-150 on MiSeq for a 254 bp sequence (46 bp overlap). This gives a high quality read similar to what we would get with a 454 read but with much more coverage. Analyzed reads with QIME. We followed a recent app note describing this.

The PGM should be able to do the same thing if their PE reads are long enough. We went with the MiSeq because we already run HiSeqs and have the library prep process down and really like Illumina's one-step automated Paired-End sequencing. In terms of cost per run the MiSeq is currently better but PGM catching up. I also prefer my output reported in >/= %Q30, I don't think the PGM's Q17 and Q20 reports show the read quality is there yet.

[Jean]

I'd run these type of samples on a MiSeq. Sequence the V4 region of 16S by amplifying using primers 515F and 806R, run libraries PE-150 on MiSeq for a 254 bp sequence (46 bp overlap). This gives a high quality read similar to what we would get with a 454 read but with much more coverage. Analyzed reads with QIME. We followed a recent app note describing this.

The PGM should be able to do the same thing if their PE reads are long enough. We went with the MiSeq because we already run HiSeqs and have the library prep process down and really like Illumina's one-step automated Paired-End sequencing. In terms of cost per run the MiSeq is currently better but PGM catching up.

I believe the OP is attempting a meta-transcriptomic experiment (mRNAseq), not a 16S profiling.

[epistatic]

Sorry to misread! I'd still prefer MiSeq. Quality reads and output is really needed. For metatranscriptomes I'd prefer the HiSeq as cost for read count per sample is even cheaper than the MiSeq.

[ramsemm]

Thanks Jean and Epistatic for the feedback!

Epistatic, yes, I am attempting a metatranscriptomic experiment, but I am also planning to do 16S and ITS profiling on the MiSeq using 515F/806R and ITS1F/ITS2 on these same samples. When you did the 16S profiling in the above mentioned way, how much PhiX did you spike in?

As for the MiSeq vs. Ion Torrent comparison, thank you both. It sounds like getting the libraries prepped for a MiSeq run would be a better bet.

Thank you both!

Anyone else out there with feedback is definitely appreciated too!