So we recently had two problems with a common solution:
#1, we hadn’t had a lab social function in awhile
#2, we wanted to try using the MinION sequencer in a “field” setting, but weren’t quite ready to try it without electricity on hand.
The obvious solution, have the entire lab over to my house for a potluck and have the students in charge of the MinION do metagenomic sequencing on chicken feces (since we have a couple ongoing chicken projects/collaborations). So Marcus Cohen and Dennett Rodriguez set out to figure out what we would need from the lab to make this work. And it turned out to be a lot! By the time we had the centrifuge, vortexers, computer, sequencer, coolers, tips, pipettes etc… it was quite the mound of stuff. We decided to use the Qiagen DNeasy Blood and Tissue Kit for DNA extraction, as we know some groups have had success with that for Nanopore sequencing. The plan was to also do some culture-based work, in particular enrichments for Salmonella since the presence or absence of Salmonella in backyard chickens is a topic I often hear discussed but never in the context of evidence.
We took a ton of pictures, and live-tweeted (sort of) the action. However, we fairly quickly ran into a host of problems that in total resulted in the termination of the experiment. Here’s a quick rundown of the issues, only some of which directly relate to the MinION per se.
1) So it doesn’t take very much chicken feces to clog an extraction filter. Had to modify the protocol on the fly
2) End result of the first DNA extraction… virtually no DNA whatsoever
3) The MinKNOW software kept trying to update itself (and failing) which kept us from QCing the flow cell
4) Once we got the software sorted out (after an hour of putzing around), the QC failed… closer inspection revealed that the brand-new flowcell was dried out and ruined. Luckily we had an old backup flowcell on hand.
5) A modified repeat DNA extraction failed. Our Twitter feeds by this point had informed us that bird poop is “notoriously difficult” to get DNA out of and that there were a host of special trick required for this to work. Bad choice of target material on our part!
So lessons learned: Don’t try a new substance/extraction protocol in the field. Check the flowcell before leaving the lab.
I really want to like the MinION… really I do. But our first experience with it was a little bit painful only to discover that there are so many errors in the data that it can’t be used for bacterial genomics (our preferred application). We still really want it to work for metagenomics where some of that error might not matter as much. Hopefully we can work out the details!