Storage For Miseq In-House

Question

[Lee Katz]

Labs with MiSeqs or other benchtop genome sequencers:

What is your in-house solution for storing the data and then serving it out to the lab? I am imagining something like a raided hard drive for the hardware and then something like galaxy to serve it out. But really, I have no clue right now what to do. What are the hardware and software solutions you have developed?

Answer 1

[Josh Herr]

The lab I work in has a relatively small server (we bought the PSSC labs: Whole Genome Sequencing Server) and storage unit (also from PSSC labs: SureStore U1000 Storage Array but we only have 300TB in it, which we have almost filled up with storing raw data runs, assembled genomes, reference genomes and databases, and transcriptomes, multiple copies of raw data in people's folders, etc., so we're probably adding more memory soon).

Seems to work for us, the people in the lab who use it typically SSH into the shell and do everything from there. The only serious limitation for the hardware is we work on some large plant genomes (and large metagenomes) and assembly memory can be the limiting factor.

[Istvan Albert]

That's the cheapest system I've seen when considering the specs! Most other vendors ask more, usually a lot more.

[Josh Herr]

We have the 16 core Intel, 256 GB RAM version which is not the cheapest version, but it's not prohibitively expensive either. I think PSSC is reputable; the customer service for setting everything up wasn't stellar and the cooling system is a little loud, but no serious complaints. It's only about a year off the shelf, which is why I haven't been over to use that awesome Fedora desktop workstation you have Istvan!