Differences in OTUs between Hiseq and Miseq runs

[Maureen Berg]

Hello,

This isn't really QIIME-specific, but I though I'd ask here anyway. I am trying to compare two experiments, one run on a Miseq (250bp PE reads) and the other on the Hiseq (150bp reads). I have compared different experiments/different Hiseq runs in the past, and the dominate OTUs were always similar between experiments and runs. However, I tried the Miseq for my latest experiment. When comparing one Hiseq experiment with the Miseq experiment (they are biological repeats of each other), I get different OTUs for each experiment. The overall trends are still the same (the composition at the family level, for example, is similar between experiments). Has anyone else experienced this?

I used closed reference OTU picking, so its not an issue of de novo OTUs.

Thanks!

Maureen

[Greg Caporaso]

Hi Maureen,

We've observed differences based on different read lengths, but not differences based on sequencing technology. Could that be what you're seeing? You could try trimming the MiSeq reads to 150 bases, and then redoing your comparison. (Note I'm assuming you're using the same primers for both runs - if not, that would definitely contribute to this.)

We'll be interested to hear what you find. 

Greg

[Maureen Berg]

Greg,

I trimmed my longer reads, and I think it fixed the issue! I just picked OTUs on a subset (~25-30%) of the data to test this out. I will re-run the entire dataset with the trimmed reads.

Thank you!

Maureen

[Greg Caporaso]

Great, thanks for following up!