We’ve sat on this one for a while as we thought we could do better next time round. Back in December 2016 we decided to trial our tools and protocols for nanopore sequencing without mains electricity or access to the internet using a lightweight assemblage of kit. We thought about various places we could try: the beach in Aberystwyth, the summit of Snowdon, the International Space Station. But then we realised we had an ideal location: somewhere we could reach in the department van rather than a rocket. A coal mine.
Although a considerable portion of Earth’s biomass lives in the subsurface, as far as we know, until now nobody has been silly enough to try DNA sequencing while still within the subsurface. To see how our kit performed, we thought we would give it a go.
We managed to extract, quantify, library prep and run the sequencer while at a depth of -100m, but we faced some challenges along the way which mean we would like to try again to improve the data. For now, because others have had an interest in offline, off-grid sequencing we thought we’d share our protocol and kit as a preprint which we’ll update in due course.
Things which worked well:
- The sequencing itself. Our MinION had no problem sucking strands through >1100 active pores while underground. MinKNOW basecalled the reads while offline too. Thanks to our miner minder Paul Green for pressing the go button on the run
- Quantification on Qubit. Sara Rassner got consistent dsDNA yields from our extracts using the Qubit while it was powered by a lithium powerpack.
- Library prep: Gloved hands and a thermos mug provided the 30 and 75 degrees Celsius incubation steps. Our MiniPCR cycler had to stay at home as the powerpack had to be hooked up to the Qubit at all times.
- The workflow. We had very limited time to spend underground. Everyone worked like greased gazelles and the process ran smoothly for the most part
- The not having a methane explosion while underground bit. Coal mines present this additional hazard. We had to take detailed precautions, risk assessments and clearances to get our kit underground. One precaution meant connecting up and running all electrical kit while aboveground – limiting our flexibility and battery life.
Things we need to work on
- DNA extraction yield and integrity from low biomass high mineral content samples. We settled on ochreous biofilms as our target substrate as a result of prior positive experience, but our yields from a conventional PowerSoil protocol were very poor. On the basis “one molecule, one read” and that we had a very happy flow cell on the MinION we decided to crack on. I put as much DNA as we could in to the library, squeezing 3x the normal library volume in to the flow cell whereever we could (SpotON port, normal port…bit more into the SpotON…etc). This still totaled <10% of the recommended input of DNA. This is the likely cause of the short read length and low percentage of identified reads. We’re working on this.
- The battery powered centrifuge. True, non-electric drill based battery powered centrifuges are thin on the ground. We found a 12V DC option which we hacked to run off a battery. It has a disconcerting habit of shattering some bead tubes every so often. Other than rescaling & pooling extracts we’re not sure how to solve this.
PhD student and caving bioinformatician André Soares was in his element running the “data mining” for this project. He made a video of the event. He’s also won a competition to attend London Calling thanks to his #deepseq tweet and I gather he’ll have a poster to present there too. Lucky fella.
Our thanks to all the staff at The Big Pit museum who made the sequencing run possible.