RNA-Seq on 8 lane vs 2 lane sequencer

[silverfugue]

Hi, I've been lurking these forums for a while, and this is my first time posting.

I'm running an RNA-Seq experiment that will use 4 lanes and will be multiplexing all of my samples on each lane. The sequencing facility that I will be sending samples to has two Illumina HiSeq machines, one with 8 lanes, and another that has 2 lanes. Would there be a difference between sequencing my samples on the 8 lane machine (all 4 lanes run on same batch), or sequencing my samples on the 2 lane machine (run 2 lanes, then run the "third and fourth" lanes on another sequencing run)?

Thanks!

[scbaker]

Running your samples on two separate machines will add a bit of technical variability, but since all of your samples are multiplexed together, my guess is that it won't have a significant impact on your results.

However, I'm not sure what you mean by an '8 lane machine' and a '2 lane machine'. The HiSeq 2500 has two modes - 'high output' (which uses flow cells with 8 lanes; up to 300Gb per flow cell, 2 flow cells per run) and 'rapid mode' (which uses flow cells with 2 lanes, up to 60Gb per flow cell, 2 flow cells per run). But it's the same machine - you just decide if you want to pay a bit extra money (on a per Gb basis) for the rapid mode, which gives you results back in about 24 hours rather than a week or more.

[GenoMax]

silverfugue perhaps was referring to a HiSeq 2000 (8 lane) and a HiSeq 2500 (2 lane).

Since the samples are multiplexed it should not matter a whole lot but if one wants to be extra cautious then sticking with the same kind of machine (8 or 2 lane) will eliminate one factor.

[silverfugue]

Thank you for your replies. The facility does have 2 machines, one in which they run mostly the HiSeq 2000 (8 lane) and another for HiSeq 2500 (2 lane), sorry if I didn't make that clear in the original post. Based on your replies, if I chose to run my "4 lanes" on the HiSeq 2500 platform twice (and on the same machine) vs on the HiSeq 2000, it shouldn't add many more additional factors for variation.

[scbaker]

Thank you for your replies. The facility does have 2 machines, one in which they run mostly the HiSeq 2000 (8 lane) and another for HiSeq 2500 (2 lane), sorry if I didn't make that clear in the original post. Based on your replies, if I chose to run my "4 lanes" on the HiSeq 2500 platform twice (and on the same machine) vs on the HiSeq 2000, it shouldn't add many more additional factors for variation.

Well, running your samples in two separate runs will add SOME variability, so I would avoid it if possible. I would recommend doing one of the following:
1) 4 lanes on a single 2000 flow cell
2) a full run on a 2500 (2 flow cells, 2 lanes each for a total of 4 lanes)

Either of these is a bit better than 2 separate runs on the 2500.